@mastra/core 0.21.1 → 0.21.2-alpha.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (212) hide show
  1. package/CHANGELOG.md +44 -0
  2. package/dist/agent/__tests__/mock-model.d.ts +6 -0
  3. package/dist/agent/__tests__/mock-model.d.ts.map +1 -0
  4. package/dist/agent/agent.d.ts +9 -4
  5. package/dist/agent/agent.d.ts.map +1 -1
  6. package/dist/agent/agent.types.d.ts +4 -3
  7. package/dist/agent/agent.types.d.ts.map +1 -1
  8. package/dist/agent/index.cjs +13 -13
  9. package/dist/agent/index.js +2 -2
  10. package/dist/agent/input-processor/index.cjs +6 -6
  11. package/dist/agent/input-processor/index.js +1 -1
  12. package/dist/agent/message-list/index.d.ts.map +1 -1
  13. package/dist/agent/message-list/utils/ai-v5/gemini-compatibility.d.ts +22 -0
  14. package/dist/agent/message-list/utils/ai-v5/gemini-compatibility.d.ts.map +1 -0
  15. package/dist/agent/types.d.ts +5 -0
  16. package/dist/agent/types.d.ts.map +1 -1
  17. package/dist/agent/utils.d.ts.map +1 -1
  18. package/dist/agent/workflows/prepare-stream/index.d.ts +7 -4
  19. package/dist/agent/workflows/prepare-stream/index.d.ts.map +1 -1
  20. package/dist/agent/workflows/prepare-stream/prepare-memory-step.d.ts +7 -4
  21. package/dist/agent/workflows/prepare-stream/prepare-memory-step.d.ts.map +1 -1
  22. package/dist/agent/workflows/prepare-stream/schema.d.ts +4 -4
  23. package/dist/agent/workflows/prepare-stream/stream-step.d.ts +3 -0
  24. package/dist/agent/workflows/prepare-stream/stream-step.d.ts.map +1 -1
  25. package/dist/ai-tracing/index.cjs +32 -32
  26. package/dist/ai-tracing/index.js +1 -1
  27. package/dist/{chunk-NHKXTGR6.js → chunk-4ZFVJA3A.js} +3 -3
  28. package/dist/{chunk-NHKXTGR6.js.map → chunk-4ZFVJA3A.js.map} +1 -1
  29. package/dist/{chunk-XVREJCOW.cjs → chunk-7P5FMQQK.cjs} +6 -6
  30. package/dist/{chunk-XVREJCOW.cjs.map → chunk-7P5FMQQK.cjs.map} +1 -1
  31. package/dist/{chunk-3V4O35FO.js → chunk-BLVB2IJK.js} +3 -3
  32. package/dist/{chunk-3V4O35FO.js.map → chunk-BLVB2IJK.js.map} +1 -1
  33. package/dist/{chunk-FESSBG5D.cjs → chunk-D3IOLHNB.cjs} +12 -12
  34. package/dist/{chunk-FESSBG5D.cjs.map → chunk-D3IOLHNB.cjs.map} +1 -1
  35. package/dist/{chunk-KAV4EP3T.cjs → chunk-EGNHIPGJ.cjs} +4 -4
  36. package/dist/{chunk-KAV4EP3T.cjs.map → chunk-EGNHIPGJ.cjs.map} +1 -1
  37. package/dist/{chunk-5NR2GY4U.js → chunk-GB5BIA4V.js} +5 -3
  38. package/dist/chunk-GB5BIA4V.js.map +1 -0
  39. package/dist/{chunk-J2OKI2L3.js → chunk-GKTRUHKB.js} +10 -5
  40. package/dist/chunk-GKTRUHKB.js.map +1 -0
  41. package/dist/{chunk-A3QHQYMC.js → chunk-HGNRQ3OG.js} +12 -2
  42. package/dist/chunk-HGNRQ3OG.js.map +1 -0
  43. package/dist/{chunk-MUAFIPD7.js → chunk-KLWDUSBX.js} +4 -4
  44. package/dist/{chunk-MUAFIPD7.js.map → chunk-KLWDUSBX.js.map} +1 -1
  45. package/dist/chunk-LJFJTTZQ.cjs +4 -0
  46. package/dist/chunk-LJFJTTZQ.cjs.map +1 -0
  47. package/dist/{chunk-T3PBU7JK.js → chunk-LMP5KMU3.js} +34 -45
  48. package/dist/chunk-LMP5KMU3.js.map +1 -0
  49. package/dist/{chunk-OXAN5SM7.cjs → chunk-LRSB62Z6.cjs} +12 -2
  50. package/dist/chunk-LRSB62Z6.cjs.map +1 -0
  51. package/dist/{chunk-3ORA5EFD.cjs → chunk-LZUSZT7R.cjs} +5 -3
  52. package/dist/chunk-LZUSZT7R.cjs.map +1 -0
  53. package/dist/{chunk-XHPBKALR.cjs → chunk-MVVMHOAC.cjs} +15 -12
  54. package/dist/chunk-MVVMHOAC.cjs.map +1 -0
  55. package/dist/{chunk-HBJVAMCS.js → chunk-N3ZSX5OS.js} +31 -32
  56. package/dist/chunk-N3ZSX5OS.js.map +1 -0
  57. package/dist/{chunk-HP7B2TQQ.js → chunk-NIESGYOH.js} +1654 -1469
  58. package/dist/chunk-NIESGYOH.js.map +1 -0
  59. package/dist/{chunk-F6BMIPYW.js → chunk-NL3P7IFI.js} +3 -3
  60. package/dist/{chunk-F6BMIPYW.js.map → chunk-NL3P7IFI.js.map} +1 -1
  61. package/dist/{chunk-U6CUPKV2.js → chunk-R35PINKN.js} +5 -5
  62. package/dist/{chunk-U6CUPKV2.js.map → chunk-R35PINKN.js.map} +1 -1
  63. package/dist/{chunk-W3DS6F6G.cjs → chunk-UN653TND.cjs} +4 -4
  64. package/dist/{chunk-W3DS6F6G.cjs.map → chunk-UN653TND.cjs.map} +1 -1
  65. package/dist/{chunk-44ODA6L3.cjs → chunk-URQI7RJ3.cjs} +4 -4
  66. package/dist/{chunk-44ODA6L3.cjs.map → chunk-URQI7RJ3.cjs.map} +1 -1
  67. package/dist/{chunk-APY2BE4K.cjs → chunk-URVM3Y4T.cjs} +35 -46
  68. package/dist/chunk-URVM3Y4T.cjs.map +1 -0
  69. package/dist/{chunk-AJK7QPSD.cjs → chunk-V5WDWSQ5.cjs} +11 -11
  70. package/dist/{chunk-AJK7QPSD.cjs.map → chunk-V5WDWSQ5.cjs.map} +1 -1
  71. package/dist/{chunk-FLXB6WXO.js → chunk-VQLW7LPO.js} +5 -3
  72. package/dist/chunk-VQLW7LPO.js.map +1 -0
  73. package/dist/{chunk-462DRETM.cjs → chunk-WI2YMBNG.cjs} +64 -66
  74. package/dist/chunk-WI2YMBNG.cjs.map +1 -0
  75. package/dist/{chunk-NUURZ3AY.js → chunk-XBJP7THY.js} +6 -3
  76. package/dist/chunk-XBJP7THY.js.map +1 -0
  77. package/dist/{chunk-5DZ4RYWL.js → chunk-XYNSPEJO.js} +3 -3
  78. package/dist/{chunk-5DZ4RYWL.js.map → chunk-XYNSPEJO.js.map} +1 -1
  79. package/dist/{chunk-OY6UJ3QL.cjs → chunk-YGL6E6PD.cjs} +5 -3
  80. package/dist/chunk-YGL6E6PD.cjs.map +1 -0
  81. package/dist/{chunk-BV6CD4IZ.cjs → chunk-YKVVW65D.cjs} +12 -7
  82. package/dist/chunk-YKVVW65D.cjs.map +1 -0
  83. package/dist/{chunk-DGQW5VRR.cjs → chunk-ZMPSVDDO.cjs} +1658 -1472
  84. package/dist/chunk-ZMPSVDDO.cjs.map +1 -0
  85. package/dist/chunk-ZV5CC35D.js +3 -0
  86. package/dist/chunk-ZV5CC35D.js.map +1 -0
  87. package/dist/index.cjs +50 -46
  88. package/dist/index.js +8 -8
  89. package/dist/llm/index.cjs +10 -6
  90. package/dist/llm/index.d.ts +1 -0
  91. package/dist/llm/index.d.ts.map +1 -1
  92. package/dist/llm/index.js +1 -1
  93. package/dist/llm/model/embedding-router.d.ts +50 -0
  94. package/dist/llm/model/embedding-router.d.ts.map +1 -0
  95. package/dist/llm/model/gateways/models-dev.d.ts.map +1 -1
  96. package/dist/llm/model/gateways/netlify.d.ts.map +1 -1
  97. package/dist/llm/model/index.d.ts +1 -0
  98. package/dist/llm/model/index.d.ts.map +1 -1
  99. package/dist/llm/model/model.loop.d.ts +1 -1
  100. package/dist/llm/model/model.loop.d.ts.map +1 -1
  101. package/dist/llm/model/provider-options.d.ts +42 -0
  102. package/dist/llm/model/provider-options.d.ts.map +1 -0
  103. package/dist/llm/model/provider-types.generated.d.ts +629 -590
  104. package/dist/llm/model/router.d.ts.map +1 -1
  105. package/dist/loop/index.cjs +2 -2
  106. package/dist/loop/index.js +1 -1
  107. package/dist/loop/network/index.d.ts +4 -1
  108. package/dist/loop/network/index.d.ts.map +1 -1
  109. package/dist/loop/test-utils/fullStream.d.ts.map +1 -1
  110. package/dist/loop/test-utils/generateText.d.ts.map +1 -1
  111. package/dist/loop/test-utils/options.d.ts.map +1 -1
  112. package/dist/loop/test-utils/resultObject.d.ts.map +1 -1
  113. package/dist/loop/test-utils/streamObject.d.ts.map +1 -1
  114. package/dist/loop/test-utils/telemetry.d.ts.map +1 -1
  115. package/dist/loop/test-utils/textStream.d.ts.map +1 -1
  116. package/dist/loop/test-utils/toUIMessageStream.d.ts.map +1 -1
  117. package/dist/loop/test-utils/tools.d.ts.map +1 -1
  118. package/dist/loop/test-utils/utils.d.ts +2 -0
  119. package/dist/loop/test-utils/utils.d.ts.map +1 -1
  120. package/dist/loop/workflows/agentic-execution/index.d.ts +42 -42
  121. package/dist/loop/workflows/agentic-execution/llm-execution-step.d.ts +28 -28
  122. package/dist/loop/workflows/agentic-execution/llm-mapping-step.d.ts +14 -14
  123. package/dist/loop/workflows/agentic-loop/index.d.ts +42 -42
  124. package/dist/loop/workflows/schema.d.ts +14 -14
  125. package/dist/mastra/index.cjs +2 -2
  126. package/dist/mastra/index.js +1 -1
  127. package/dist/memory/index.cjs +4 -4
  128. package/dist/memory/index.js +1 -1
  129. package/dist/memory/memory.d.ts.map +1 -1
  130. package/dist/memory/types.d.ts +10 -1
  131. package/dist/memory/types.d.ts.map +1 -1
  132. package/dist/models-dev-4VGIWYS3.js +3 -0
  133. package/dist/{models-dev-3VEABUH4.js.map → models-dev-4VGIWYS3.js.map} +1 -1
  134. package/dist/models-dev-AXZASLL2.cjs +12 -0
  135. package/dist/{models-dev-L42ORGVT.cjs.map → models-dev-AXZASLL2.cjs.map} +1 -1
  136. package/dist/netlify-26MMY5X4.js +3 -0
  137. package/dist/{netlify-NYFKA6UB.js.map → netlify-26MMY5X4.js.map} +1 -1
  138. package/dist/netlify-F3MVGV4Y.cjs +12 -0
  139. package/dist/{netlify-3B6HDHRP.cjs.map → netlify-F3MVGV4Y.cjs.map} +1 -1
  140. package/dist/processors/index.cjs +11 -11
  141. package/dist/processors/index.js +1 -1
  142. package/dist/provider-registry.json +1542 -1464
  143. package/dist/relevance/index.cjs +4 -4
  144. package/dist/relevance/index.js +1 -1
  145. package/dist/scores/index.cjs +9 -9
  146. package/dist/scores/index.js +2 -2
  147. package/dist/scores/scoreTraces/index.cjs +8 -8
  148. package/dist/scores/scoreTraces/index.js +3 -3
  149. package/dist/storage/domains/workflows/inmemory.d.ts.map +1 -1
  150. package/dist/storage/index.cjs +5 -3
  151. package/dist/storage/index.cjs.map +1 -1
  152. package/dist/storage/index.js +3 -1
  153. package/dist/storage/index.js.map +1 -1
  154. package/dist/stream/aisdk/v5/execute.d.ts.map +1 -1
  155. package/dist/stream/aisdk/v5/output.d.ts +9 -10
  156. package/dist/stream/aisdk/v5/output.d.ts.map +1 -1
  157. package/dist/stream/base/output-format-handlers.d.ts +1 -2
  158. package/dist/stream/base/output-format-handlers.d.ts.map +1 -1
  159. package/dist/stream/base/output.d.ts.map +1 -1
  160. package/dist/stream/index.cjs +10 -10
  161. package/dist/stream/index.d.ts +1 -1
  162. package/dist/stream/index.d.ts.map +1 -1
  163. package/dist/stream/index.js +2 -2
  164. package/dist/stream/types.d.ts +17 -1
  165. package/dist/stream/types.d.ts.map +1 -1
  166. package/dist/test-utils/llm-mock.cjs +2 -2
  167. package/dist/test-utils/llm-mock.js +1 -1
  168. package/dist/tools/index.cjs +2 -2
  169. package/dist/tools/index.js +1 -1
  170. package/dist/tools/stream.d.ts +7 -0
  171. package/dist/tools/stream.d.ts.map +1 -1
  172. package/dist/utils.cjs +17 -17
  173. package/dist/utils.js +1 -1
  174. package/dist/workflows/default.d.ts +1 -0
  175. package/dist/workflows/default.d.ts.map +1 -1
  176. package/dist/workflows/evented/index.cjs +10 -10
  177. package/dist/workflows/evented/index.js +1 -1
  178. package/dist/workflows/evented/workflow-event-processor/index.d.ts.map +1 -1
  179. package/dist/workflows/evented/workflow.d.ts.map +1 -1
  180. package/dist/workflows/index.cjs +12 -12
  181. package/dist/workflows/index.js +1 -1
  182. package/dist/workflows/legacy/index.cjs +22 -22
  183. package/dist/workflows/legacy/index.js +1 -1
  184. package/dist/workflows/legacy/machine.d.ts +1 -1
  185. package/dist/workflows/step.d.ts +4 -1
  186. package/dist/workflows/step.d.ts.map +1 -1
  187. package/dist/workflows/types.d.ts +1 -0
  188. package/dist/workflows/types.d.ts.map +1 -1
  189. package/dist/workflows/workflow.d.ts +3 -1
  190. package/dist/workflows/workflow.d.ts.map +1 -1
  191. package/package.json +11 -10
  192. package/src/llm/model/provider-types.generated.d.ts +629 -590
  193. package/dist/chunk-3ORA5EFD.cjs.map +0 -1
  194. package/dist/chunk-462DRETM.cjs.map +0 -1
  195. package/dist/chunk-5NR2GY4U.js.map +0 -1
  196. package/dist/chunk-A3QHQYMC.js.map +0 -1
  197. package/dist/chunk-APY2BE4K.cjs.map +0 -1
  198. package/dist/chunk-BV6CD4IZ.cjs.map +0 -1
  199. package/dist/chunk-DGQW5VRR.cjs.map +0 -1
  200. package/dist/chunk-FLXB6WXO.js.map +0 -1
  201. package/dist/chunk-HBJVAMCS.js.map +0 -1
  202. package/dist/chunk-HP7B2TQQ.js.map +0 -1
  203. package/dist/chunk-J2OKI2L3.js.map +0 -1
  204. package/dist/chunk-NUURZ3AY.js.map +0 -1
  205. package/dist/chunk-OXAN5SM7.cjs.map +0 -1
  206. package/dist/chunk-OY6UJ3QL.cjs.map +0 -1
  207. package/dist/chunk-T3PBU7JK.js.map +0 -1
  208. package/dist/chunk-XHPBKALR.cjs.map +0 -1
  209. package/dist/models-dev-3VEABUH4.js +0 -3
  210. package/dist/models-dev-L42ORGVT.cjs +0 -12
  211. package/dist/netlify-3B6HDHRP.cjs +0 -12
  212. package/dist/netlify-NYFKA6UB.js +0 -3
@@ -1,118 +1,101 @@
1
- import { ModelsDevGateway, parseModelRouterId } from './chunk-5NR2GY4U.js';
2
- import { NetlifyGateway } from './chunk-FLXB6WXO.js';
1
+ import { ModelsDevGateway, parseModelRouterId } from './chunk-GB5BIA4V.js';
2
+ import { NetlifyGateway } from './chunk-VQLW7LPO.js';
3
3
  import { RuntimeContext } from './chunk-HLRWYUFN.js';
4
4
  import { MastraError } from './chunk-T3JFFQH2.js';
5
- import { createHash } from 'crypto';
6
- import { createOpenAICompatible } from '@ai-sdk/openai-compatible-v5';
7
5
  import fs from 'fs';
8
6
  import { createRequire } from 'module';
9
7
  import os from 'os';
10
8
  import path from 'path';
11
-
12
- // src/llm/model/gateways/index.ts
13
- function findGatewayForModel(gatewayId, gateways2) {
14
- const prefixedGateway = gateways2.find((g) => g.prefix && gatewayId.startsWith(`${g.prefix}/`));
15
- if (prefixedGateway) {
16
- return prefixedGateway;
17
- }
18
- const unprefixedGateways = gateways2.filter((g) => !g.prefix);
19
- for (const gateway of unprefixedGateways) {
20
- return gateway;
21
- }
22
- throw new MastraError({
23
- id: "MODEL_ROUTER_NO_GATEWAY_FOUND",
24
- category: "USER",
25
- domain: "MODEL_ROUTER",
26
- text: `No Mastra model router gateway found for model id ${gatewayId}`
27
- });
28
- }
9
+ import { createHash } from 'crypto';
10
+ import { createOpenAICompatible } from '@ai-sdk/openai-compatible-v5';
11
+ import { createGoogleGenerativeAI } from '@ai-sdk/google-v5';
12
+ import { createOpenAI } from '@ai-sdk/openai-v5';
29
13
 
30
14
  // src/llm/model/provider-registry.json
31
15
  var provider_registry_default = {
32
16
  providers: {
33
- "moonshotai-cn": {
34
- url: "https://api.moonshot.cn/v1",
35
- apiKeyEnvVar: "MOONSHOT_API_KEY",
36
- apiKeyHeader: "Authorization",
37
- name: "Moonshot AI (China)",
38
- models: ["kimi-k2-0711-preview", "kimi-k2-0905-preview", "kimi-k2-turbo-preview"],
39
- docUrl: "https://platform.moonshot.cn/docs/api/chat",
40
- gateway: "models.dev"
41
- },
42
- lucidquery: {
43
- url: "https://lucidquery.com/api/v1",
44
- apiKeyEnvVar: "LUCIDQUERY_API_KEY",
17
+ nvidia: {
18
+ url: "https://integrate.api.nvidia.com/v1",
19
+ apiKeyEnvVar: "NVIDIA_API_KEY",
45
20
  apiKeyHeader: "Authorization",
46
- name: "LucidQuery AI",
47
- models: ["lucidnova-rf1-100b", "lucidquery-nexus-coder"],
48
- docUrl: "https://lucidquery.com/api/docs",
21
+ name: "Nvidia",
22
+ models: [
23
+ "black-forest-labs/flux.1-dev",
24
+ "deepseek-ai/deepseek-v3.1",
25
+ "deepseek-ai/deepseek-v3.1-terminus",
26
+ "google/gemma-3-27b-it",
27
+ "microsoft/phi-4-mini-instruct",
28
+ "moonshotai/kimi-k2-instruct",
29
+ "moonshotai/kimi-k2-instruct-0905",
30
+ "nvidia/cosmos-nemotron-34b",
31
+ "nvidia/llama-3.1-nemotron-ultra-253b-v1",
32
+ "nvidia/nemoretriever-ocr-v1",
33
+ "nvidia/parakeet-tdt-0.6b-v2",
34
+ "openai/gpt-oss-120b",
35
+ "openai/whisper-large-v3",
36
+ "qwen/qwen3-235b-a22b",
37
+ "qwen/qwen3-coder-480b-a35b-instruct"
38
+ ],
39
+ docUrl: "https://docs.api.nvidia.com/nim/",
49
40
  gateway: "models.dev"
50
41
  },
51
- moonshotai: {
52
- url: "https://api.moonshot.ai/v1",
53
- apiKeyEnvVar: "MOONSHOT_API_KEY",
42
+ modelscope: {
43
+ url: "https://api-inference.modelscope.cn/v1",
44
+ apiKeyEnvVar: "MODELSCOPE_API_KEY",
54
45
  apiKeyHeader: "Authorization",
55
- name: "Moonshot AI",
56
- models: ["kimi-k2-0711-preview", "kimi-k2-0905-preview", "kimi-k2-turbo-preview"],
57
- docUrl: "https://platform.moonshot.ai/docs/api/chat",
46
+ name: "ModelScope",
47
+ models: [
48
+ "Qwen/Qwen3-235B-A22B-Instruct-2507",
49
+ "Qwen/Qwen3-235B-A22B-Thinking-2507",
50
+ "Qwen/Qwen3-30B-A3B-Instruct-2507",
51
+ "Qwen/Qwen3-30B-A3B-Thinking-2507",
52
+ "Qwen/Qwen3-Coder-30B-A3B-Instruct",
53
+ "ZhipuAI/GLM-4.5",
54
+ "ZhipuAI/GLM-4.6"
55
+ ],
56
+ docUrl: "https://modelscope.cn/docs/model-service/API-Inference/intro",
58
57
  gateway: "models.dev"
59
58
  },
60
- "zai-coding-plan": {
61
- url: "https://api.z.ai/api/coding/paas/v4",
62
- apiKeyEnvVar: "ZHIPU_API_KEY",
59
+ aihubmix: {
60
+ url: "https://api.aihubmix.com/v1",
61
+ apiKeyEnvVar: "AIHUBMIX_API_KEY",
63
62
  apiKeyHeader: "Authorization",
64
- name: "Z.AI Coding Plan",
65
- models: ["glm-4.5", "glm-4.5-air", "glm-4.5-flash", "glm-4.5v", "glm-4.6"],
66
- docUrl: "https://docs.z.ai/devpack/overview",
63
+ name: "AIHubMix",
64
+ models: [
65
+ "DeepSeek-V3.2-Exp",
66
+ "DeepSeek-V3.2-Exp-Think",
67
+ "Kimi-K2-0905",
68
+ "claude-haiku-4-5",
69
+ "claude-opus-4-1",
70
+ "claude-sonnet-4-5",
71
+ "gemini-2.5-flash",
72
+ "gemini-2.5-pro",
73
+ "glm-4.6",
74
+ "gpt-4.1",
75
+ "gpt-4.1-mini",
76
+ "gpt-4.1-nano",
77
+ "gpt-4o",
78
+ "gpt-4o-2024-11-20",
79
+ "gpt-5",
80
+ "gpt-5-codex",
81
+ "gpt-5-mini",
82
+ "gpt-5-nano",
83
+ "gpt-5-pro",
84
+ "o4-mini",
85
+ "qwen3-235b-a22b-instruct-2507",
86
+ "qwen3-235b-a22b-thinking-2507",
87
+ "qwen3-coder-480b-a35b-instruct"
88
+ ],
89
+ docUrl: "https://aihubmix.com/docs/models",
67
90
  gateway: "models.dev"
68
91
  },
69
- alibaba: {
70
- url: "https://dashscope-intl.aliyuncs.com/compatible-mode/v1",
71
- apiKeyEnvVar: "DASHSCOPE_API_KEY",
92
+ lucidquery: {
93
+ url: "https://lucidquery.com/api/v1",
94
+ apiKeyEnvVar: "LUCIDQUERY_API_KEY",
72
95
  apiKeyHeader: "Authorization",
73
- name: "Alibaba",
74
- models: [
75
- "qvq-max",
76
- "qwen-flash",
77
- "qwen-max",
78
- "qwen-mt-plus",
79
- "qwen-mt-turbo",
80
- "qwen-omni-turbo",
81
- "qwen-omni-turbo-realtime",
82
- "qwen-plus",
83
- "qwen-plus-character-ja",
84
- "qwen-turbo",
85
- "qwen-vl-max",
86
- "qwen-vl-ocr",
87
- "qwen-vl-plus",
88
- "qwen2-5-14b-instruct",
89
- "qwen2-5-32b-instruct",
90
- "qwen2-5-72b-instruct",
91
- "qwen2-5-7b-instruct",
92
- "qwen2-5-omni-7b",
93
- "qwen2-5-vl-72b-instruct",
94
- "qwen2-5-vl-7b-instruct",
95
- "qwen3-14b",
96
- "qwen3-235b-a22b",
97
- "qwen3-32b",
98
- "qwen3-8b",
99
- "qwen3-asr-flash",
100
- "qwen3-coder-30b-a3b-instruct",
101
- "qwen3-coder-480b-a35b-instruct",
102
- "qwen3-coder-flash",
103
- "qwen3-coder-plus",
104
- "qwen3-livetranslate-flash-realtime",
105
- "qwen3-max",
106
- "qwen3-next-80b-a3b-instruct",
107
- "qwen3-next-80b-a3b-thinking",
108
- "qwen3-omni-flash",
109
- "qwen3-omni-flash-realtime",
110
- "qwen3-vl-235b-a22b",
111
- "qwen3-vl-30b-a3b",
112
- "qwen3-vl-plus",
113
- "qwq-plus"
114
- ],
115
- docUrl: "https://www.alibabacloud.com/help/en/model-studio/models",
96
+ name: "LucidQuery AI",
97
+ models: ["lucidnova-rf1-100b", "lucidquery-nexus-coder"],
98
+ docUrl: "https://lucidquery.com/api/docs",
116
99
  gateway: "models.dev"
117
100
  },
118
101
  xai: {
@@ -143,537 +126,450 @@ var provider_registry_default = {
143
126
  docUrl: "https://docs.x.ai/docs/models",
144
127
  gateway: "models.dev"
145
128
  },
146
- nvidia: {
147
- url: "https://integrate.api.nvidia.com/v1",
148
- apiKeyEnvVar: "NVIDIA_API_KEY",
129
+ deepseek: {
130
+ url: "https://api.deepseek.com",
131
+ apiKeyEnvVar: "DEEPSEEK_API_KEY",
149
132
  apiKeyHeader: "Authorization",
150
- name: "Nvidia",
151
- models: [
152
- "black-forest-labs/flux.1-dev",
153
- "deepseek-ai/deepseek-v3.1",
154
- "google/gemma-3-27b-it",
155
- "microsoft/phi-4-mini-instruct",
156
- "moonshotai/kimi-k2-instruct",
157
- "moonshotai/kimi-k2-instruct-0905",
158
- "nvidia/cosmos-nemotron-34b",
159
- "nvidia/llama-3.1-nemotron-ultra-253b-v1",
160
- "nvidia/nemoretriever-ocr-v1",
161
- "nvidia/parakeet-tdt-0.6b-v2",
162
- "openai/gpt-oss-120b",
163
- "openai/whisper-large-v3",
164
- "qwen/qwen3-235b-a22b",
165
- "qwen/qwen3-coder-480b-a35b-instruct"
166
- ],
167
- docUrl: "https://docs.api.nvidia.com/nim/",
133
+ name: "DeepSeek",
134
+ models: ["deepseek-chat", "deepseek-reasoner"],
135
+ docUrl: "https://platform.deepseek.com/api-docs/pricing",
168
136
  gateway: "models.dev"
169
137
  },
170
- upstage: {
171
- url: "https://api.upstage.ai",
172
- apiKeyEnvVar: "UPSTAGE_API_KEY",
138
+ requesty: {
139
+ url: "https://router.requesty.ai/v1",
140
+ apiKeyEnvVar: "REQUESTY_API_KEY",
173
141
  apiKeyHeader: "Authorization",
174
- name: "Upstage",
175
- models: ["solar-mini", "solar-pro2"],
176
- docUrl: "https://developers.upstage.ai/docs/apis/chat",
142
+ name: "Requesty",
143
+ models: [
144
+ "anthropic/claude-3-7-sonnet",
145
+ "anthropic/claude-4-sonnet-20250522",
146
+ "anthropic/claude-opus-4",
147
+ "anthropic/claude-opus-4-1-20250805",
148
+ "google/gemini-2.5-flash",
149
+ "google/gemini-2.5-pro",
150
+ "openai/gpt-4.1",
151
+ "openai/gpt-4.1-mini",
152
+ "openai/gpt-4o-mini",
153
+ "openai/gpt-5",
154
+ "openai/gpt-5-mini",
155
+ "openai/gpt-5-nano",
156
+ "openai/o4-mini"
157
+ ],
158
+ docUrl: "https://requesty.ai/solution/llm-routing/models",
177
159
  gateway: "models.dev"
178
160
  },
179
- groq: {
180
- url: "https://api.groq.com/openai/v1",
181
- apiKeyEnvVar: "GROQ_API_KEY",
161
+ nebius: {
162
+ url: "https://api.studio.nebius.com/v1/",
163
+ apiKeyEnvVar: "NEBIUS_API_KEY",
182
164
  apiKeyHeader: "Authorization",
183
- name: "Groq",
165
+ name: "Nebius AI Studio",
184
166
  models: [
185
- "deepseek-r1-distill-llama-70b",
186
- "gemma2-9b-it",
187
- "llama-3.1-8b-instant",
188
- "llama-3.3-70b-versatile",
189
- "llama-guard-3-8b",
190
- "llama3-70b-8192",
191
- "llama3-8b-8192",
192
- "meta-llama/llama-4-maverick-17b-128e-instruct",
193
- "meta-llama/llama-4-scout-17b-16e-instruct",
194
- "meta-llama/llama-guard-4-12b",
195
- "mistral-saba-24b",
167
+ "NousResearch/hermes-4-405b",
168
+ "NousResearch/hermes-4-70b",
169
+ "deepseek-ai/deepseek-v3",
170
+ "meta-llama/llama-3.3-70b-instruct-base",
171
+ "meta-llama/llama-3.3-70b-instruct-fast",
172
+ "meta-llama/llama-3_1-405b-instruct",
196
173
  "moonshotai/kimi-k2-instruct",
197
- "moonshotai/kimi-k2-instruct-0905",
174
+ "nvidia/llama-3_1-nemotron-ultra-253b-v1",
198
175
  "openai/gpt-oss-120b",
199
176
  "openai/gpt-oss-20b",
200
- "qwen-qwq-32b",
201
- "qwen/qwen3-32b"
177
+ "qwen/qwen3-235b-a22b-instruct-2507",
178
+ "qwen/qwen3-235b-a22b-thinking-2507",
179
+ "qwen/qwen3-coder-480b-a35b-instruct",
180
+ "zai-org/glm-4.5",
181
+ "zai-org/glm-4.5-air"
202
182
  ],
203
- docUrl: "https://console.groq.com/docs/models",
183
+ docUrl: "https://docs.studio.nebius.com/quickstart",
204
184
  gateway: "models.dev"
205
185
  },
206
- mistral: {
207
- url: "https://api.mistral.ai/v1",
208
- apiKeyEnvVar: "MISTRAL_API_KEY",
186
+ fastrouter: {
187
+ url: "https://go.fastrouter.ai/api/v1",
188
+ apiKeyEnvVar: "FASTROUTER_API_KEY",
209
189
  apiKeyHeader: "Authorization",
210
- name: "Mistral",
190
+ name: "FastRouter",
211
191
  models: [
212
- "codestral-latest",
213
- "devstral-medium-2507",
214
- "devstral-small-2505",
215
- "devstral-small-2507",
216
- "magistral-medium-latest",
217
- "magistral-small",
218
- "ministral-3b-latest",
219
- "ministral-8b-latest",
220
- "mistral-large-latest",
221
- "mistral-medium-2505",
222
- "mistral-medium-2508",
223
- "mistral-medium-latest",
224
- "mistral-nemo",
225
- "mistral-small-latest",
226
- "open-mistral-7b",
227
- "open-mixtral-8x22b",
228
- "open-mixtral-8x7b",
229
- "pixtral-12b",
230
- "pixtral-large-latest"
192
+ "anthropic/claude-opus-4.1",
193
+ "anthropic/claude-sonnet-4",
194
+ "deepseek-ai/deepseek-r1-distill-llama-70b",
195
+ "google/gemini-2.5-flash",
196
+ "google/gemini-2.5-pro",
197
+ "moonshotai/kimi-k2",
198
+ "openai/gpt-4.1",
199
+ "openai/gpt-5",
200
+ "openai/gpt-5-mini",
201
+ "openai/gpt-5-nano",
202
+ "openai/gpt-oss-120b",
203
+ "openai/gpt-oss-20b",
204
+ "qwen/qwen3-coder",
205
+ "x-ai/grok-4"
231
206
  ],
232
- docUrl: "https://docs.mistral.ai/getting-started/models/",
207
+ docUrl: "https://fastrouter.ai/models",
233
208
  gateway: "models.dev"
234
209
  },
235
- vercel: {
236
- url: "https://ai-gateway.vercel.sh/v1",
237
- apiKeyEnvVar: "AI_GATEWAY_API_KEY",
210
+ perplexity: {
211
+ url: "https://api.perplexity.ai",
212
+ apiKeyEnvVar: "PERPLEXITY_API_KEY",
238
213
  apiKeyHeader: "Authorization",
239
- name: "Vercel AI Gateway",
214
+ name: "Perplexity",
215
+ models: ["sonar", "sonar-pro", "sonar-reasoning", "sonar-reasoning-pro"],
216
+ docUrl: "https://docs.perplexity.ai",
217
+ gateway: "models.dev"
218
+ },
219
+ morph: {
220
+ url: "https://api.morphllm.com/v1",
221
+ apiKeyEnvVar: "MORPH_API_KEY",
222
+ apiKeyHeader: "Authorization",
223
+ name: "Morph",
224
+ models: ["auto", "morph-v3-fast", "morph-v3-large"],
225
+ docUrl: "https://docs.morphllm.com/api-reference/introduction",
226
+ gateway: "models.dev"
227
+ },
228
+ "zai-coding-plan": {
229
+ url: "https://api.z.ai/api/coding/paas/v4",
230
+ apiKeyEnvVar: "ZHIPU_API_KEY",
231
+ apiKeyHeader: "Authorization",
232
+ name: "Z.AI Coding Plan",
233
+ models: ["glm-4.5", "glm-4.5-air", "glm-4.5-flash", "glm-4.5v", "glm-4.6"],
234
+ docUrl: "https://docs.z.ai/devpack/overview",
235
+ gateway: "models.dev"
236
+ },
237
+ huggingface: {
238
+ url: "https://router.huggingface.co/v1",
239
+ apiKeyEnvVar: "HF_TOKEN",
240
+ apiKeyHeader: "Authorization",
241
+ name: "Hugging Face",
240
242
  models: [
241
- "alibaba/qwen3-coder-plus",
242
- "alibaba/qwen3-max",
243
- "alibaba/qwen3-next-80b-a3b-instruct",
244
- "alibaba/qwen3-next-80b-a3b-thinking",
245
- "alibaba/qwen3-vl-instruct",
246
- "alibaba/qwen3-vl-thinking",
247
- "amazon/nova-lite",
248
- "amazon/nova-micro",
249
- "amazon/nova-pro",
250
- "anthropic/claude-3-5-haiku",
251
- "anthropic/claude-3-haiku",
252
- "anthropic/claude-3-opus",
253
- "anthropic/claude-3.5-sonnet",
243
+ "Qwen/Qwen3-235B-A22B-Thinking-2507",
244
+ "Qwen/Qwen3-Coder-480B-A35B-Instruct",
245
+ "Qwen/Qwen3-Next-80B-A3B-Instruct",
246
+ "Qwen/Qwen3-Next-80B-A3B-Thinking",
247
+ "deepseek-ai/DeepSeek-R1-0528",
248
+ "deepseek-ai/Deepseek-V3-0324",
249
+ "moonshotai/Kimi-K2-Instruct",
250
+ "moonshotai/Kimi-K2-Instruct-0905",
251
+ "zai-org/GLM-4.5",
252
+ "zai-org/GLM-4.5-Air",
253
+ "zai-org/GLM-4.6"
254
+ ],
255
+ docUrl: "https://huggingface.co/docs/inference-providers",
256
+ gateway: "models.dev"
257
+ },
258
+ openrouter: {
259
+ url: "https://openrouter.ai/api/v1",
260
+ apiKeyEnvVar: "OPENROUTER_API_KEY",
261
+ name: "OpenRouter",
262
+ models: [
263
+ "anthropic/claude-3.5-haiku",
254
264
  "anthropic/claude-3.7-sonnet",
255
- "anthropic/claude-4-1-opus",
256
- "anthropic/claude-4-opus",
257
- "anthropic/claude-4-sonnet",
258
- "anthropic/claude-4.5-sonnet",
259
265
  "anthropic/claude-haiku-4.5",
260
- "cerebras/qwen3-coder",
261
- "deepseek/deepseek-r1",
266
+ "anthropic/claude-opus-4",
267
+ "anthropic/claude-opus-4.1",
268
+ "anthropic/claude-sonnet-4",
269
+ "anthropic/claude-sonnet-4.5",
270
+ "cognitivecomputations/dolphin3.0-mistral-24b",
271
+ "cognitivecomputations/dolphin3.0-r1-mistral-24b",
272
+ "deepseek/deepseek-chat-v3-0324",
273
+ "deepseek/deepseek-chat-v3.1",
274
+ "deepseek/deepseek-r1-0528-qwen3-8b:free",
275
+ "deepseek/deepseek-r1-0528:free",
262
276
  "deepseek/deepseek-r1-distill-llama-70b",
277
+ "deepseek/deepseek-r1-distill-qwen-14b",
278
+ "deepseek/deepseek-r1:free",
279
+ "deepseek/deepseek-v3-base:free",
263
280
  "deepseek/deepseek-v3.1-terminus",
264
- "deepseek/deepseek-v3.2-exp",
265
- "deepseek/deepseek-v3.2-exp-thinking",
266
- "google/gemini-2.0-flash",
267
- "google/gemini-2.0-flash-lite",
281
+ "featherless/qwerky-72b",
282
+ "google/gemini-2.0-flash-001",
283
+ "google/gemini-2.0-flash-exp:free",
268
284
  "google/gemini-2.5-flash",
269
285
  "google/gemini-2.5-flash-lite",
270
286
  "google/gemini-2.5-flash-lite-preview-09-2025",
271
287
  "google/gemini-2.5-flash-preview-09-2025",
272
288
  "google/gemini-2.5-pro",
273
- "meta/llama-3.3-70b",
274
- "meta/llama-4-maverick",
275
- "meta/llama-4-scout",
276
- "mistral/codestral",
277
- "mistral/magistral-medium",
278
- "mistral/magistral-small",
279
- "mistral/ministral-3b",
280
- "mistral/ministral-8b",
281
- "mistral/mistral-large",
282
- "mistral/mistral-small",
283
- "mistral/mixtral-8x22b-instruct",
284
- "mistral/pixtral-12b",
285
- "mistral/pixtral-large",
289
+ "google/gemini-2.5-pro-preview-05-06",
290
+ "google/gemini-2.5-pro-preview-06-05",
291
+ "google/gemma-2-9b-it:free",
292
+ "google/gemma-3-12b-it",
293
+ "google/gemma-3-27b-it",
294
+ "google/gemma-3n-e4b-it",
295
+ "google/gemma-3n-e4b-it:free",
296
+ "meta-llama/llama-3.2-11b-vision-instruct",
297
+ "meta-llama/llama-3.3-70b-instruct:free",
298
+ "meta-llama/llama-4-scout:free",
299
+ "microsoft/mai-ds-r1:free",
300
+ "mistralai/codestral-2508",
301
+ "mistralai/devstral-medium-2507",
302
+ "mistralai/devstral-small-2505",
303
+ "mistralai/devstral-small-2505:free",
304
+ "mistralai/devstral-small-2507",
305
+ "mistralai/mistral-7b-instruct:free",
306
+ "mistralai/mistral-medium-3",
307
+ "mistralai/mistral-medium-3.1",
308
+ "mistralai/mistral-nemo:free",
309
+ "mistralai/mistral-small-3.1-24b-instruct",
310
+ "mistralai/mistral-small-3.2-24b-instruct",
311
+ "mistralai/mistral-small-3.2-24b-instruct:free",
312
+ "moonshotai/kimi-dev-72b:free",
286
313
  "moonshotai/kimi-k2",
287
- "morph/morph-v3-fast",
288
- "morph/morph-v3-large",
289
- "openai/gpt-4-turbo",
314
+ "moonshotai/kimi-k2-0905",
315
+ "moonshotai/kimi-k2:free",
316
+ "nousresearch/deephermes-3-llama-3-8b-preview",
317
+ "nousresearch/hermes-4-405b",
318
+ "nousresearch/hermes-4-70b",
290
319
  "openai/gpt-4.1",
291
320
  "openai/gpt-4.1-mini",
292
- "openai/gpt-4.1-nano",
293
- "openai/gpt-4o",
294
321
  "openai/gpt-4o-mini",
295
322
  "openai/gpt-5",
323
+ "openai/gpt-5-chat",
296
324
  "openai/gpt-5-codex",
325
+ "openai/gpt-5-image",
297
326
  "openai/gpt-5-mini",
298
327
  "openai/gpt-5-nano",
299
328
  "openai/gpt-oss-120b",
300
329
  "openai/gpt-oss-20b",
301
- "openai/o1",
302
- "openai/o3",
303
- "openai/o3-mini",
304
330
  "openai/o4-mini",
305
- "perplexity/sonar",
306
- "perplexity/sonar-pro",
307
- "perplexity/sonar-reasoning",
308
- "perplexity/sonar-reasoning-pro",
309
- "vercel/v0-1.0-md",
310
- "vercel/v0-1.5-md",
311
- "xai/grok-2",
312
- "xai/grok-2-vision",
313
- "xai/grok-3",
314
- "xai/grok-3-fast",
315
- "xai/grok-3-mini",
316
- "xai/grok-3-mini-fast",
317
- "xai/grok-4",
318
- "xai/grok-4-fast",
319
- "xai/grok-4-fast-non-reasoning",
320
- "xai/grok-code-fast-1",
321
- "zai/glm-4.5",
322
- "zai/glm-4.5-air",
323
- "zai/glm-4.5v",
324
- "zai/glm-4.6"
331
+ "openrouter/cypher-alpha:free",
332
+ "openrouter/horizon-alpha",
333
+ "openrouter/horizon-beta",
334
+ "openrouter/sonoma-dusk-alpha",
335
+ "openrouter/sonoma-sky-alpha",
336
+ "qwen/qwen-2.5-coder-32b-instruct",
337
+ "qwen/qwen2.5-vl-32b-instruct:free",
338
+ "qwen/qwen2.5-vl-72b-instruct",
339
+ "qwen/qwen2.5-vl-72b-instruct:free",
340
+ "qwen/qwen3-14b:free",
341
+ "qwen/qwen3-235b-a22b-07-25",
342
+ "qwen/qwen3-235b-a22b-07-25:free",
343
+ "qwen/qwen3-235b-a22b-thinking-2507",
344
+ "qwen/qwen3-235b-a22b:free",
345
+ "qwen/qwen3-30b-a3b-instruct-2507",
346
+ "qwen/qwen3-30b-a3b-thinking-2507",
347
+ "qwen/qwen3-30b-a3b:free",
348
+ "qwen/qwen3-32b:free",
349
+ "qwen/qwen3-8b:free",
350
+ "qwen/qwen3-coder",
351
+ "qwen/qwen3-coder:free",
352
+ "qwen/qwen3-max",
353
+ "qwen/qwen3-next-80b-a3b-instruct",
354
+ "qwen/qwen3-next-80b-a3b-thinking",
355
+ "qwen/qwq-32b:free",
356
+ "rekaai/reka-flash-3",
357
+ "sarvamai/sarvam-m:free",
358
+ "thudm/glm-z1-32b:free",
359
+ "tngtech/deepseek-r1t2-chimera:free",
360
+ "x-ai/grok-3",
361
+ "x-ai/grok-3-beta",
362
+ "x-ai/grok-3-mini",
363
+ "x-ai/grok-3-mini-beta",
364
+ "x-ai/grok-4",
365
+ "x-ai/grok-4-fast",
366
+ "x-ai/grok-4-fast:free",
367
+ "x-ai/grok-code-fast-1",
368
+ "z-ai/glm-4.5",
369
+ "z-ai/glm-4.5-air",
370
+ "z-ai/glm-4.5-air:free",
371
+ "z-ai/glm-4.5v",
372
+ "z-ai/glm-4.6"
325
373
  ],
326
- docUrl: "https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway",
374
+ docUrl: "https://openrouter.ai/models",
327
375
  gateway: "models.dev"
328
376
  },
329
- nebius: {
330
- url: "https://api.studio.nebius.com/v1/",
331
- apiKeyEnvVar: "NEBIUS_API_KEY",
377
+ cerebras: {
378
+ url: "https://api.cerebras.ai/v1",
379
+ apiKeyEnvVar: "CEREBRAS_API_KEY",
332
380
  apiKeyHeader: "Authorization",
333
- name: "Nebius AI Studio",
334
- models: [
335
- "NousResearch/hermes-4-405b",
336
- "NousResearch/hermes-4-70b",
337
- "deepseek-ai/deepseek-v3",
338
- "meta-llama/llama-3.3-70b-instruct-base",
339
- "meta-llama/llama-3.3-70b-instruct-fast",
340
- "meta-llama/llama-3_1-405b-instruct",
341
- "moonshotai/kimi-k2-instruct",
342
- "nvidia/llama-3_1-nemotron-ultra-253b-v1",
343
- "openai/gpt-oss-120b",
344
- "openai/gpt-oss-20b",
345
- "qwen/qwen3-235b-a22b-instruct-2507",
346
- "qwen/qwen3-235b-a22b-thinking-2507",
347
- "qwen/qwen3-coder-480b-a35b-instruct",
348
- "zai-org/glm-4.5",
349
- "zai-org/glm-4.5-air"
350
- ],
351
- docUrl: "https://docs.studio.nebius.com/quickstart",
381
+ name: "Cerebras",
382
+ models: ["gpt-oss-120b", "qwen-3-235b-a22b-instruct-2507", "qwen-3-coder-480b"],
383
+ docUrl: "https://inference-docs.cerebras.ai/models/overview",
352
384
  gateway: "models.dev"
353
385
  },
354
- deepseek: {
355
- url: "https://api.deepseek.com",
356
- apiKeyEnvVar: "DEEPSEEK_API_KEY",
386
+ zhipuai: {
387
+ url: "https://open.bigmodel.cn/api/paas/v4",
388
+ apiKeyEnvVar: "ZHIPU_API_KEY",
357
389
  apiKeyHeader: "Authorization",
358
- name: "DeepSeek",
359
- models: ["deepseek-chat", "deepseek-reasoner"],
360
- docUrl: "https://platform.deepseek.com/api-docs/pricing",
390
+ name: "Zhipu AI",
391
+ models: ["glm-4.5", "glm-4.5-air", "glm-4.5-flash", "glm-4.5v", "glm-4.6"],
392
+ docUrl: "https://docs.z.ai/guides/overview/pricing",
361
393
  gateway: "models.dev"
362
394
  },
363
- "alibaba-cn": {
364
- url: "https://dashscope.aliyuncs.com/compatible-mode/v1",
365
- apiKeyEnvVar: "DASHSCOPE_API_KEY",
395
+ togetherai: {
396
+ url: "https://api.together.xyz/v1",
397
+ apiKeyEnvVar: "TOGETHER_API_KEY",
366
398
  apiKeyHeader: "Authorization",
367
- name: "Alibaba (China)",
399
+ name: "Together AI",
368
400
  models: [
369
- "deepseek-r1",
370
- "deepseek-r1-0528",
371
- "deepseek-r1-distill-llama-70b",
372
- "deepseek-r1-distill-llama-8b",
373
- "deepseek-r1-distill-qwen-1-5b",
374
- "deepseek-r1-distill-qwen-14b",
375
- "deepseek-r1-distill-qwen-32b",
376
- "deepseek-r1-distill-qwen-7b",
377
- "deepseek-v3",
378
- "deepseek-v3-1",
379
- "deepseek-v3-2-exp",
380
- "moonshot-kimi-k2-instruct",
381
- "qvq-max",
382
- "qwen-deep-research",
383
- "qwen-doc-turbo",
384
- "qwen-flash",
385
- "qwen-long",
386
- "qwen-math-plus",
387
- "qwen-math-turbo",
388
- "qwen-max",
389
- "qwen-mt-plus",
390
- "qwen-mt-turbo",
391
- "qwen-omni-turbo",
392
- "qwen-omni-turbo-realtime",
393
- "qwen-plus",
394
- "qwen-plus-character",
395
- "qwen-turbo",
396
- "qwen-vl-max",
397
- "qwen-vl-ocr",
398
- "qwen-vl-plus",
399
- "qwen2-5-14b-instruct",
400
- "qwen2-5-32b-instruct",
401
- "qwen2-5-72b-instruct",
402
- "qwen2-5-7b-instruct",
403
- "qwen2-5-coder-32b-instruct",
404
- "qwen2-5-coder-7b-instruct",
405
- "qwen2-5-math-72b-instruct",
406
- "qwen2-5-math-7b-instruct",
407
- "qwen2-5-omni-7b",
408
- "qwen2-5-vl-72b-instruct",
409
- "qwen2-5-vl-7b-instruct",
410
- "qwen3-14b",
411
- "qwen3-235b-a22b",
412
- "qwen3-32b",
413
- "qwen3-8b",
414
- "qwen3-asr-flash",
415
- "qwen3-coder-30b-a3b-instruct",
416
- "qwen3-coder-480b-a35b-instruct",
417
- "qwen3-coder-flash",
418
- "qwen3-coder-plus",
419
- "qwen3-max",
420
- "qwen3-next-80b-a3b-instruct",
421
- "qwen3-next-80b-a3b-thinking",
422
- "qwen3-omni-flash",
423
- "qwen3-omni-flash-realtime",
424
- "qwen3-vl-235b-a22b",
425
- "qwen3-vl-30b-a3b",
426
- "qwen3-vl-plus",
427
- "qwq-32b",
428
- "qwq-plus",
429
- "tongyi-intent-detect-v3"
401
+ "Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8",
402
+ "deepseek-ai/DeepSeek-R1",
403
+ "deepseek-ai/DeepSeek-V3",
404
+ "meta-llama/Llama-3.3-70B-Instruct-Turbo",
405
+ "moonshotai/Kimi-K2-Instruct",
406
+ "openai/gpt-oss-120b"
430
407
  ],
431
- docUrl: "https://www.alibabacloud.com/help/en/model-studio/models",
408
+ docUrl: "https://docs.together.ai/docs/serverless-models",
432
409
  gateway: "models.dev"
433
410
  },
434
- venice: {
435
- url: "https://api.venice.ai/api/v1",
436
- apiKeyEnvVar: "VENICE_API_KEY",
411
+ wandb: {
412
+ url: "https://api.inference.wandb.ai/v1",
413
+ apiKeyEnvVar: "WANDB_API_KEY",
437
414
  apiKeyHeader: "Authorization",
438
- name: "Venice AI",
415
+ name: "Weights & Biases",
439
416
  models: [
440
- "deepseek-coder-v2-lite",
441
- "deepseek-r1-671b",
442
- "dolphin-2.9.2-qwen2-72b",
443
- "llama-3.1-405b",
444
- "llama-3.2-3b",
445
- "llama-3.3-70b",
446
- "mistral-31-24b",
447
- "qwen-2.5-coder-32b",
448
- "qwen-2.5-qwq-32b",
449
- "qwen-2.5-vl",
450
- "qwen3-235b",
451
- "qwen3-4b",
452
- "venice-uncensored"
417
+ "Qwen/Qwen3-235B-A22B-Instruct-2507",
418
+ "Qwen/Qwen3-235B-A22B-Thinking-2507",
419
+ "Qwen/Qwen3-Coder-480B-A35B-Instruct",
420
+ "deepseek-ai/DeepSeek-R1-0528",
421
+ "deepseek-ai/DeepSeek-V3-0324",
422
+ "meta-llama/Llama-3.1-8B-Instruct",
423
+ "meta-llama/Llama-3.3-70B-Instruct",
424
+ "meta-llama/Llama-4-Scout-17B-16E-Instruct",
425
+ "microsoft/Phi-4-mini-instruct",
426
+ "moonshotai/Kimi-K2-Instruct"
453
427
  ],
454
- docUrl: "https://docs.venice.ai",
428
+ docUrl: "https://weave-docs.wandb.ai/guides/integrations/inference/",
455
429
  gateway: "models.dev"
456
430
  },
457
- chutes: {
458
- url: "https://llm.chutes.ai/v1",
459
- apiKeyEnvVar: "CHUTES_API_KEY",
431
+ submodel: {
432
+ url: "https://llm.submodel.ai/v1",
433
+ apiKeyEnvVar: "SUBMODEL_INSTAGEN_ACCESS_KEY",
460
434
  apiKeyHeader: "Authorization",
461
- name: "Chutes",
435
+ name: "submodel",
462
436
  models: [
463
437
  "Qwen/Qwen3-235B-A22B-Instruct-2507",
464
438
  "Qwen/Qwen3-235B-A22B-Thinking-2507",
465
- "Qwen/Qwen3-30B-A3B",
466
- "Qwen/Qwen3-30B-A3B-Instruct-2507",
467
- "Qwen/Qwen3-30B-A3B-Thinking-2507",
468
- "Qwen/Qwen3-Coder-30B-A3B-Instruct",
469
439
  "Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8",
470
- "Qwen/Qwen3-Next-80B-A3B-Instruct",
471
- "Qwen/Qwen3-Next-80B-A3B-Thinking",
472
- "chutesai/Devstral-Small-2505",
473
- "chutesai/Mistral-Small-3.2-24B-Instruct-2506",
474
440
  "deepseek-ai/DeepSeek-R1-0528",
475
- "deepseek-ai/DeepSeek-R1-0528-Qwen3-8B",
476
- "deepseek-ai/DeepSeek-R1-Distill-Llama-70B",
477
441
  "deepseek-ai/DeepSeek-V3-0324",
478
442
  "deepseek-ai/DeepSeek-V3.1",
479
- "deepseek-ai/DeepSeek-V3.1-Terminus",
480
- "deepseek-ai/DeepSeek-V3.1-turbo",
481
- "deepseek-ai/DeepSeek-V3.1:THINKING",
482
- "deepseek-ai/DeepSeek-V3.2-Exp",
483
- "meituan-longcat/LongCat-Flash-Chat-FP8",
484
- "moonshotai/Kimi-Dev-72B",
485
- "moonshotai/Kimi-K2-Instruct-0905",
486
- "moonshotai/Kimi-K2-Instruct-75k",
487
- "moonshotai/Kimi-VL-A3B-Thinking",
488
443
  "openai/gpt-oss-120b",
489
- "tngtech/DeepSeek-R1T-Chimera",
490
- "tngtech/DeepSeek-TNG-R1T2-Chimera",
491
444
  "zai-org/GLM-4.5-Air",
492
- "zai-org/GLM-4.5-FP8",
493
- "zai-org/GLM-4.5-turbo",
494
- "zai-org/GLM-4.6-FP8",
495
- "zai-org/GLM-4.6-turbo"
496
- ],
497
- docUrl: "https://llm.chutes.ai/v1/models",
498
- gateway: "models.dev"
499
- },
500
- cortecs: {
501
- url: "https://api.cortecs.ai/v1",
502
- apiKeyEnvVar: "CORTECS_API_KEY",
503
- apiKeyHeader: "Authorization",
504
- name: "Cortecs",
505
- models: [
506
- "claude-4-5-sonnet",
507
- "claude-sonnet-4",
508
- "deepseek-v3-0324",
509
- "gemini-2.5-pro",
510
- "gpt-4.1",
511
- "gpt-oss-120b",
512
- "kimi-k2-instruct",
513
- "llama-3.1-405b-instruct",
514
- "nova-pro-v1",
515
- "qwen3-32b",
516
- "qwen3-coder-480b-a35b-instruct"
445
+ "zai-org/GLM-4.5-FP8"
517
446
  ],
518
- docUrl: "https://api.cortecs.ai/v1/models",
447
+ docUrl: "https://submodel.gitbook.io",
519
448
  gateway: "models.dev"
520
449
  },
521
- "github-models": {
522
- url: "https://models.github.ai/inference",
523
- apiKeyEnvVar: "GITHUB_TOKEN",
450
+ vercel: {
451
+ url: "https://ai-gateway.vercel.sh/v1",
452
+ apiKeyEnvVar: "AI_GATEWAY_API_KEY",
524
453
  apiKeyHeader: "Authorization",
525
- name: "GitHub Models",
454
+ name: "Vercel AI Gateway",
526
455
  models: [
527
- "ai21-labs/ai21-jamba-1.5-large",
528
- "ai21-labs/ai21-jamba-1.5-mini",
529
- "cohere/cohere-command-a",
530
- "cohere/cohere-command-r",
531
- "cohere/cohere-command-r-08-2024",
532
- "cohere/cohere-command-r-plus",
533
- "cohere/cohere-command-r-plus-08-2024",
534
- "core42/jais-30b-chat",
456
+ "alibaba/qwen3-coder-plus",
457
+ "alibaba/qwen3-max",
458
+ "alibaba/qwen3-next-80b-a3b-instruct",
459
+ "alibaba/qwen3-next-80b-a3b-thinking",
460
+ "alibaba/qwen3-vl-instruct",
461
+ "alibaba/qwen3-vl-thinking",
462
+ "amazon/nova-lite",
463
+ "amazon/nova-micro",
464
+ "amazon/nova-pro",
465
+ "anthropic/claude-3-5-haiku",
466
+ "anthropic/claude-3-haiku",
467
+ "anthropic/claude-3-opus",
468
+ "anthropic/claude-3.5-sonnet",
469
+ "anthropic/claude-3.7-sonnet",
470
+ "anthropic/claude-4-1-opus",
471
+ "anthropic/claude-4-opus",
472
+ "anthropic/claude-4-sonnet",
473
+ "anthropic/claude-4.5-sonnet",
474
+ "anthropic/claude-haiku-4.5",
475
+ "cerebras/qwen3-coder",
535
476
  "deepseek/deepseek-r1",
536
- "deepseek/deepseek-r1-0528",
537
- "deepseek/deepseek-v3-0324",
538
- "meta/llama-3.2-11b-vision-instruct",
539
- "meta/llama-3.2-90b-vision-instruct",
540
- "meta/llama-3.3-70b-instruct",
541
- "meta/llama-4-maverick-17b-128e-instruct-fp8",
542
- "meta/llama-4-scout-17b-16e-instruct",
543
- "meta/meta-llama-3-70b-instruct",
544
- "meta/meta-llama-3-8b-instruct",
545
- "meta/meta-llama-3.1-405b-instruct",
546
- "meta/meta-llama-3.1-70b-instruct",
547
- "meta/meta-llama-3.1-8b-instruct",
548
- "microsoft/mai-ds-r1",
549
- "microsoft/phi-3-medium-128k-instruct",
550
- "microsoft/phi-3-medium-4k-instruct",
551
- "microsoft/phi-3-mini-128k-instruct",
552
- "microsoft/phi-3-mini-4k-instruct",
553
- "microsoft/phi-3-small-128k-instruct",
554
- "microsoft/phi-3-small-8k-instruct",
555
- "microsoft/phi-3.5-mini-instruct",
556
- "microsoft/phi-3.5-moe-instruct",
557
- "microsoft/phi-3.5-vision-instruct",
558
- "microsoft/phi-4",
559
- "microsoft/phi-4-mini-instruct",
560
- "microsoft/phi-4-mini-reasoning",
561
- "microsoft/phi-4-multimodal-instruct",
562
- "microsoft/phi-4-reasoning",
563
- "mistral-ai/codestral-2501",
564
- "mistral-ai/ministral-3b",
565
- "mistral-ai/mistral-large-2411",
566
- "mistral-ai/mistral-medium-2505",
567
- "mistral-ai/mistral-nemo",
568
- "mistral-ai/mistral-small-2503",
477
+ "deepseek/deepseek-r1-distill-llama-70b",
478
+ "deepseek/deepseek-v3.1-terminus",
479
+ "deepseek/deepseek-v3.2-exp",
480
+ "deepseek/deepseek-v3.2-exp-thinking",
481
+ "google/gemini-2.0-flash",
482
+ "google/gemini-2.0-flash-lite",
483
+ "google/gemini-2.5-flash",
484
+ "google/gemini-2.5-flash-lite",
485
+ "google/gemini-2.5-flash-lite-preview-09-2025",
486
+ "google/gemini-2.5-flash-preview-09-2025",
487
+ "google/gemini-2.5-pro",
488
+ "meta/llama-3.3-70b",
489
+ "meta/llama-4-maverick",
490
+ "meta/llama-4-scout",
491
+ "mistral/codestral",
492
+ "mistral/magistral-medium",
493
+ "mistral/magistral-small",
494
+ "mistral/ministral-3b",
495
+ "mistral/ministral-8b",
496
+ "mistral/mistral-large",
497
+ "mistral/mistral-small",
498
+ "mistral/mixtral-8x22b-instruct",
499
+ "mistral/pixtral-12b",
500
+ "mistral/pixtral-large",
501
+ "moonshotai/kimi-k2",
502
+ "morph/morph-v3-fast",
503
+ "morph/morph-v3-large",
504
+ "openai/gpt-4-turbo",
569
505
  "openai/gpt-4.1",
570
506
  "openai/gpt-4.1-mini",
571
507
  "openai/gpt-4.1-nano",
572
508
  "openai/gpt-4o",
573
509
  "openai/gpt-4o-mini",
510
+ "openai/gpt-5",
511
+ "openai/gpt-5-codex",
512
+ "openai/gpt-5-mini",
513
+ "openai/gpt-5-nano",
514
+ "openai/gpt-oss-120b",
515
+ "openai/gpt-oss-20b",
574
516
  "openai/o1",
575
- "openai/o1-mini",
576
- "openai/o1-preview",
577
517
  "openai/o3",
578
518
  "openai/o3-mini",
579
519
  "openai/o4-mini",
520
+ "perplexity/sonar",
521
+ "perplexity/sonar-pro",
522
+ "perplexity/sonar-reasoning",
523
+ "perplexity/sonar-reasoning-pro",
524
+ "vercel/v0-1.0-md",
525
+ "vercel/v0-1.5-md",
526
+ "xai/grok-2",
527
+ "xai/grok-2-vision",
580
528
  "xai/grok-3",
581
- "xai/grok-3-mini"
582
- ],
583
- docUrl: "https://docs.github.com/en/github-models",
584
- gateway: "models.dev"
585
- },
586
- togetherai: {
587
- url: "https://api.together.xyz/v1",
588
- apiKeyEnvVar: "TOGETHER_API_KEY",
589
- apiKeyHeader: "Authorization",
590
- name: "Together AI",
591
- models: [
592
- "Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8",
593
- "deepseek-ai/DeepSeek-R1",
594
- "deepseek-ai/DeepSeek-V3",
595
- "meta-llama/Llama-3.3-70B-Instruct-Turbo",
596
- "moonshotai/Kimi-K2-Instruct",
597
- "openai/gpt-oss-120b"
598
- ],
599
- docUrl: "https://docs.together.ai/docs/serverless-models",
600
- gateway: "models.dev"
601
- },
602
- baseten: {
603
- url: "https://inference.baseten.co/v1",
604
- apiKeyEnvVar: "BASETEN_API_KEY",
605
- apiKeyHeader: "Authorization",
606
- name: "Baseten",
607
- models: ["Qwen3/Qwen3-Coder-480B-A35B-Instruct", "moonshotai/Kimi-K2-Instruct-0905"],
608
- docUrl: "https://docs.baseten.co/development/model-apis/overview",
609
- gateway: "models.dev"
610
- },
611
- huggingface: {
612
- url: "https://router.huggingface.co/v1",
613
- apiKeyEnvVar: "HF_TOKEN",
614
- apiKeyHeader: "Authorization",
615
- name: "Hugging Face",
616
- models: [
617
- "Qwen/Qwen3-235B-A22B-Thinking-2507",
618
- "Qwen/Qwen3-Coder-480B-A35B-Instruct",
619
- "Qwen/Qwen3-Next-80B-A3B-Instruct",
620
- "Qwen/Qwen3-Next-80B-A3B-Thinking",
621
- "deepseek-ai/DeepSeek-R1-0528",
622
- "deepseek-ai/Deepseek-V3-0324",
623
- "moonshotai/Kimi-K2-Instruct",
624
- "moonshotai/Kimi-K2-Instruct-0905",
625
- "zai-org/GLM-4.5",
626
- "zai-org/GLM-4.5-Air",
627
- "zai-org/GLM-4.6"
628
- ],
629
- docUrl: "https://huggingface.co/docs/inference-providers",
630
- gateway: "models.dev"
631
- },
632
- opencode: {
633
- url: "https://opencode.ai/zen/v1",
634
- apiKeyEnvVar: "OPENCODE_API_KEY",
635
- apiKeyHeader: "Authorization",
636
- name: "OpenCode Zen",
637
- models: [
638
- "an-g8x",
639
- "claude-3-5-haiku",
640
- "claude-haiku-4-5",
641
- "claude-opus-4-1",
642
- "claude-sonnet-4",
643
- "claude-sonnet-4-5",
644
- "code-supernova",
645
- "glm-4.6",
646
- "gpt-5",
647
- "gpt-5-codex",
648
- "grok-code",
649
- "kimi-k2",
650
- "qwen3-coder"
529
+ "xai/grok-3-fast",
530
+ "xai/grok-3-mini",
531
+ "xai/grok-3-mini-fast",
532
+ "xai/grok-4",
533
+ "xai/grok-4-fast",
534
+ "xai/grok-4-fast-non-reasoning",
535
+ "xai/grok-code-fast-1",
536
+ "zai/glm-4.5",
537
+ "zai/glm-4.5-air",
538
+ "zai/glm-4.5v",
539
+ "zai/glm-4.6"
651
540
  ],
652
- docUrl: "https://opencode.ai/docs/zen",
541
+ docUrl: "https://github.com/vercel/ai/tree/5eb85cc45a259553501f535b8ac79a77d0e79223/packages/gateway",
653
542
  gateway: "models.dev"
654
543
  },
655
- fastrouter: {
656
- url: "https://go.fastrouter.ai/api/v1",
657
- apiKeyEnvVar: "FASTROUTER_API_KEY",
544
+ synthetic: {
545
+ url: "https://api.synthetic.new/v1",
546
+ apiKeyEnvVar: "SYNTHETIC_API_KEY",
658
547
  apiKeyHeader: "Authorization",
659
- name: "FastRouter",
548
+ name: "Synthetic",
660
549
  models: [
661
- "anthropic/claude-opus-4.1",
662
- "anthropic/claude-sonnet-4",
663
- "deepseek-ai/deepseek-r1-distill-llama-70b",
664
- "google/gemini-2.5-flash",
665
- "google/gemini-2.5-pro",
666
- "moonshotai/kimi-k2",
667
- "openai/gpt-4.1",
668
- "openai/gpt-5",
669
- "openai/gpt-5-mini",
670
- "openai/gpt-5-nano",
671
- "openai/gpt-oss-120b",
672
- "openai/gpt-oss-20b",
673
- "qwen/qwen3-coder",
674
- "x-ai/grok-4"
550
+ "hf:Qwen/Qwen2.5-Coder-32B-Instruct",
551
+ "hf:Qwen/Qwen3-235B-A22B-Instruct-2507",
552
+ "hf:Qwen/Qwen3-235B-A22B-Thinking-2507",
553
+ "hf:Qwen/Qwen3-Coder-480B-A35B-Instruct",
554
+ "hf:deepseek-ai/DeepSeek-R1",
555
+ "hf:deepseek-ai/DeepSeek-R1-0528",
556
+ "hf:deepseek-ai/DeepSeek-V3",
557
+ "hf:deepseek-ai/DeepSeek-V3-0324",
558
+ "hf:deepseek-ai/DeepSeek-V3.1",
559
+ "hf:deepseek-ai/DeepSeek-V3.1-Terminus",
560
+ "hf:meta-llama/Llama-3.1-405B-Instruct",
561
+ "hf:meta-llama/Llama-3.1-70B-Instruct",
562
+ "hf:meta-llama/Llama-3.1-8B-Instruct",
563
+ "hf:meta-llama/Llama-3.3-70B-Instruct",
564
+ "hf:meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8",
565
+ "hf:meta-llama/Llama-4-Scout-17B-16E-Instruct",
566
+ "hf:moonshotai/Kimi-K2-Instruct",
567
+ "hf:moonshotai/Kimi-K2-Instruct-0905",
568
+ "hf:openai/gpt-oss-120b",
569
+ "hf:zai-org/GLM-4.5",
570
+ "hf:zai-org/GLM-4.6"
675
571
  ],
676
- docUrl: "https://fastrouter.ai/models",
572
+ docUrl: "https://synthetic.new/pricing",
677
573
  gateway: "models.dev"
678
574
  },
679
575
  google: {
@@ -707,236 +603,105 @@ var provider_registry_default = {
707
603
  docUrl: "https://ai.google.dev/gemini-api/docs/pricing",
708
604
  gateway: "models.dev"
709
605
  },
710
- inception: {
711
- url: "https://api.inceptionlabs.ai/v1/",
712
- apiKeyEnvVar: "INCEPTION_API_KEY",
606
+ groq: {
607
+ url: "https://api.groq.com/openai/v1",
608
+ apiKeyEnvVar: "GROQ_API_KEY",
713
609
  apiKeyHeader: "Authorization",
714
- name: "Inception",
715
- models: ["mercury", "mercury-coder"],
716
- docUrl: "https://platform.inceptionlabs.ai/docs",
610
+ name: "Groq",
611
+ models: [
612
+ "deepseek-r1-distill-llama-70b",
613
+ "gemma2-9b-it",
614
+ "llama-3.1-8b-instant",
615
+ "llama-3.3-70b-versatile",
616
+ "llama-guard-3-8b",
617
+ "llama3-70b-8192",
618
+ "llama3-8b-8192",
619
+ "meta-llama/llama-4-maverick-17b-128e-instruct",
620
+ "meta-llama/llama-4-scout-17b-16e-instruct",
621
+ "meta-llama/llama-guard-4-12b",
622
+ "mistral-saba-24b",
623
+ "moonshotai/kimi-k2-instruct",
624
+ "moonshotai/kimi-k2-instruct-0905",
625
+ "openai/gpt-oss-120b",
626
+ "openai/gpt-oss-20b",
627
+ "qwen-qwq-32b",
628
+ "qwen/qwen3-32b"
629
+ ],
630
+ docUrl: "https://console.groq.com/docs/models",
717
631
  gateway: "models.dev"
718
632
  },
719
- wandb: {
720
- url: "https://api.inference.wandb.ai/v1",
721
- apiKeyEnvVar: "WANDB_API_KEY",
633
+ "moonshotai-cn": {
634
+ url: "https://api.moonshot.cn/v1",
635
+ apiKeyEnvVar: "MOONSHOT_API_KEY",
722
636
  apiKeyHeader: "Authorization",
723
- name: "Weights & Biases",
724
- models: [
725
- "Qwen/Qwen3-235B-A22B-Instruct-2507",
726
- "Qwen/Qwen3-235B-A22B-Thinking-2507",
727
- "Qwen/Qwen3-Coder-480B-A35B-Instruct",
728
- "deepseek-ai/DeepSeek-R1-0528",
729
- "deepseek-ai/DeepSeek-V3-0324",
730
- "meta-llama/Llama-3.1-8B-Instruct",
731
- "meta-llama/Llama-3.3-70B-Instruct",
732
- "meta-llama/Llama-4-Scout-17B-16E-Instruct",
733
- "microsoft/Phi-4-mini-instruct",
734
- "moonshotai/Kimi-K2-Instruct"
735
- ],
736
- docUrl: "https://weave-docs.wandb.ai/guides/integrations/inference/",
637
+ name: "Moonshot AI (China)",
638
+ models: ["kimi-k2-0711-preview", "kimi-k2-0905-preview", "kimi-k2-turbo-preview"],
639
+ docUrl: "https://platform.moonshot.cn/docs/api/chat",
737
640
  gateway: "models.dev"
738
641
  },
739
- openai: {
740
- apiKeyEnvVar: "OPENAI_API_KEY",
741
- name: "OpenAI",
642
+ "github-models": {
643
+ url: "https://models.github.ai/inference",
644
+ apiKeyEnvVar: "GITHUB_TOKEN",
645
+ apiKeyHeader: "Authorization",
646
+ name: "GitHub Models",
742
647
  models: [
743
- "codex-mini-latest",
744
- "gpt-3.5-turbo",
745
- "gpt-4",
746
- "gpt-4-turbo",
747
- "gpt-4.1",
748
- "gpt-4.1-mini",
749
- "gpt-4.1-nano",
750
- "gpt-4o",
751
- "gpt-4o-2024-05-13",
752
- "gpt-4o-2024-08-06",
753
- "gpt-4o-2024-11-20",
754
- "gpt-4o-mini",
755
- "gpt-5",
756
- "gpt-5-chat-latest",
757
- "gpt-5-codex",
758
- "gpt-5-mini",
759
- "gpt-5-nano",
760
- "o1",
761
- "o1-mini",
762
- "o1-preview",
763
- "o1-pro",
764
- "o3",
765
- "o3-deep-research",
766
- "o3-mini",
767
- "o3-pro",
768
- "o4-mini",
769
- "o4-mini-deep-research"
770
- ],
771
- docUrl: "https://platform.openai.com/docs/models",
772
- gateway: "models.dev"
773
- },
774
- "zhipuai-coding-plan": {
775
- url: "https://open.bigmodel.cn/api/coding/paas/v4",
776
- apiKeyEnvVar: "ZHIPU_API_KEY",
777
- apiKeyHeader: "Authorization",
778
- name: "Zhipu AI Coding Plan",
779
- models: ["glm-4.5", "glm-4.5-air", "glm-4.5-flash", "glm-4.5v", "glm-4.6"],
780
- docUrl: "https://docs.bigmodel.cn/cn/coding-plan/overview",
781
- gateway: "models.dev"
782
- },
783
- perplexity: {
784
- url: "https://api.perplexity.ai",
785
- apiKeyEnvVar: "PERPLEXITY_API_KEY",
786
- apiKeyHeader: "Authorization",
787
- name: "Perplexity",
788
- models: ["sonar", "sonar-pro", "sonar-reasoning", "sonar-reasoning-pro"],
789
- docUrl: "https://docs.perplexity.ai",
790
- gateway: "models.dev"
791
- },
792
- openrouter: {
793
- url: "https://openrouter.ai/api/v1",
794
- apiKeyEnvVar: "OPENROUTER_API_KEY",
795
- name: "OpenRouter",
796
- models: [
797
- "anthropic/claude-3.5-haiku",
798
- "anthropic/claude-3.7-sonnet",
799
- "anthropic/claude-4.5-haiku",
800
- "anthropic/claude-opus-4",
801
- "anthropic/claude-opus-4.1",
802
- "anthropic/claude-sonnet-4",
803
- "anthropic/claude-sonnet-4.5",
804
- "cognitivecomputations/dolphin3.0-mistral-24b",
805
- "cognitivecomputations/dolphin3.0-r1-mistral-24b",
806
- "deepseek/deepseek-chat-v3-0324",
807
- "deepseek/deepseek-chat-v3.1",
808
- "deepseek/deepseek-r1-0528-qwen3-8b:free",
809
- "deepseek/deepseek-r1-0528:free",
810
- "deepseek/deepseek-r1-distill-llama-70b",
811
- "deepseek/deepseek-r1-distill-qwen-14b",
812
- "deepseek/deepseek-r1:free",
813
- "deepseek/deepseek-v3-base:free",
814
- "deepseek/deepseek-v3.1-terminus",
815
- "featherless/qwerky-72b",
816
- "google/gemini-2.0-flash-001",
817
- "google/gemini-2.0-flash-exp:free",
818
- "google/gemini-2.5-flash",
819
- "google/gemini-2.5-flash-lite",
820
- "google/gemini-2.5-flash-lite-preview-09-2025",
821
- "google/gemini-2.5-flash-preview-09-2025",
822
- "google/gemini-2.5-pro",
823
- "google/gemini-2.5-pro-preview-05-06",
824
- "google/gemini-2.5-pro-preview-06-05",
825
- "google/gemma-2-9b-it:free",
826
- "google/gemma-3-12b-it",
827
- "google/gemma-3-27b-it",
828
- "google/gemma-3n-e4b-it",
829
- "google/gemma-3n-e4b-it:free",
830
- "meta-llama/llama-3.2-11b-vision-instruct",
831
- "meta-llama/llama-3.3-70b-instruct:free",
832
- "meta-llama/llama-4-scout:free",
833
- "microsoft/mai-ds-r1:free",
834
- "mistralai/codestral-2508",
835
- "mistralai/devstral-medium-2507",
836
- "mistralai/devstral-small-2505",
837
- "mistralai/devstral-small-2505:free",
838
- "mistralai/devstral-small-2507",
839
- "mistralai/mistral-7b-instruct:free",
840
- "mistralai/mistral-medium-3",
841
- "mistralai/mistral-medium-3.1",
842
- "mistralai/mistral-nemo:free",
843
- "mistralai/mistral-small-3.1-24b-instruct",
844
- "mistralai/mistral-small-3.2-24b-instruct",
845
- "mistralai/mistral-small-3.2-24b-instruct:free",
846
- "moonshotai/kimi-dev-72b:free",
847
- "moonshotai/kimi-k2",
848
- "moonshotai/kimi-k2-0905",
849
- "moonshotai/kimi-k2:free",
850
- "nousresearch/deephermes-3-llama-3-8b-preview",
851
- "nousresearch/hermes-4-405b",
852
- "nousresearch/hermes-4-70b",
648
+ "ai21-labs/ai21-jamba-1.5-large",
649
+ "ai21-labs/ai21-jamba-1.5-mini",
650
+ "cohere/cohere-command-a",
651
+ "cohere/cohere-command-r",
652
+ "cohere/cohere-command-r-08-2024",
653
+ "cohere/cohere-command-r-plus",
654
+ "cohere/cohere-command-r-plus-08-2024",
655
+ "core42/jais-30b-chat",
656
+ "deepseek/deepseek-r1",
657
+ "deepseek/deepseek-r1-0528",
658
+ "deepseek/deepseek-v3-0324",
659
+ "meta/llama-3.2-11b-vision-instruct",
660
+ "meta/llama-3.2-90b-vision-instruct",
661
+ "meta/llama-3.3-70b-instruct",
662
+ "meta/llama-4-maverick-17b-128e-instruct-fp8",
663
+ "meta/llama-4-scout-17b-16e-instruct",
664
+ "meta/meta-llama-3-70b-instruct",
665
+ "meta/meta-llama-3-8b-instruct",
666
+ "meta/meta-llama-3.1-405b-instruct",
667
+ "meta/meta-llama-3.1-70b-instruct",
668
+ "meta/meta-llama-3.1-8b-instruct",
669
+ "microsoft/mai-ds-r1",
670
+ "microsoft/phi-3-medium-128k-instruct",
671
+ "microsoft/phi-3-medium-4k-instruct",
672
+ "microsoft/phi-3-mini-128k-instruct",
673
+ "microsoft/phi-3-mini-4k-instruct",
674
+ "microsoft/phi-3-small-128k-instruct",
675
+ "microsoft/phi-3-small-8k-instruct",
676
+ "microsoft/phi-3.5-mini-instruct",
677
+ "microsoft/phi-3.5-moe-instruct",
678
+ "microsoft/phi-3.5-vision-instruct",
679
+ "microsoft/phi-4",
680
+ "microsoft/phi-4-mini-instruct",
681
+ "microsoft/phi-4-mini-reasoning",
682
+ "microsoft/phi-4-multimodal-instruct",
683
+ "microsoft/phi-4-reasoning",
684
+ "mistral-ai/codestral-2501",
685
+ "mistral-ai/ministral-3b",
686
+ "mistral-ai/mistral-large-2411",
687
+ "mistral-ai/mistral-medium-2505",
688
+ "mistral-ai/mistral-nemo",
689
+ "mistral-ai/mistral-small-2503",
853
690
  "openai/gpt-4.1",
854
691
  "openai/gpt-4.1-mini",
692
+ "openai/gpt-4.1-nano",
693
+ "openai/gpt-4o",
855
694
  "openai/gpt-4o-mini",
856
- "openai/gpt-5",
857
- "openai/gpt-5-chat",
858
- "openai/gpt-5-codex",
859
- "openai/gpt-5-image",
860
- "openai/gpt-5-mini",
861
- "openai/gpt-5-nano",
862
- "openai/gpt-oss-120b",
863
- "openai/gpt-oss-20b",
695
+ "openai/o1",
696
+ "openai/o1-mini",
697
+ "openai/o1-preview",
698
+ "openai/o3",
699
+ "openai/o3-mini",
864
700
  "openai/o4-mini",
865
- "openrouter/cypher-alpha:free",
866
- "openrouter/horizon-alpha",
867
- "openrouter/horizon-beta",
868
- "openrouter/sonoma-dusk-alpha",
869
- "openrouter/sonoma-sky-alpha",
870
- "qwen/qwen-2.5-coder-32b-instruct",
871
- "qwen/qwen2.5-vl-32b-instruct:free",
872
- "qwen/qwen2.5-vl-72b-instruct",
873
- "qwen/qwen2.5-vl-72b-instruct:free",
874
- "qwen/qwen3-14b:free",
875
- "qwen/qwen3-235b-a22b-07-25",
876
- "qwen/qwen3-235b-a22b-07-25:free",
877
- "qwen/qwen3-235b-a22b-thinking-2507",
878
- "qwen/qwen3-235b-a22b:free",
879
- "qwen/qwen3-30b-a3b-instruct-2507",
880
- "qwen/qwen3-30b-a3b-thinking-2507",
881
- "qwen/qwen3-30b-a3b:free",
882
- "qwen/qwen3-32b:free",
883
- "qwen/qwen3-8b:free",
884
- "qwen/qwen3-coder",
885
- "qwen/qwen3-coder:free",
886
- "qwen/qwen3-max",
887
- "qwen/qwen3-next-80b-a3b-instruct",
888
- "qwen/qwen3-next-80b-a3b-thinking",
889
- "qwen/qwq-32b:free",
890
- "rekaai/reka-flash-3",
891
- "sarvamai/sarvam-m:free",
892
- "thudm/glm-z1-32b:free",
893
- "tngtech/deepseek-r1t2-chimera:free",
894
- "x-ai/grok-3",
895
- "x-ai/grok-3-beta",
896
- "x-ai/grok-3-mini",
897
- "x-ai/grok-3-mini-beta",
898
- "x-ai/grok-4",
899
- "x-ai/grok-4-fast",
900
- "x-ai/grok-4-fast:free",
901
- "x-ai/grok-code-fast-1",
902
- "z-ai/glm-4.5",
903
- "z-ai/glm-4.5-air",
904
- "z-ai/glm-4.5-air:free",
905
- "z-ai/glm-4.5v",
906
- "z-ai/glm-4.6"
907
- ],
908
- docUrl: "https://openrouter.ai/models",
909
- gateway: "models.dev"
910
- },
911
- synthetic: {
912
- url: "https://api.synthetic.new/v1",
913
- apiKeyEnvVar: "SYNTHETIC_API_KEY",
914
- apiKeyHeader: "Authorization",
915
- name: "Synthetic",
916
- models: [
917
- "hf:Qwen/Qwen2.5-Coder-32B-Instruct",
918
- "hf:Qwen/Qwen3-235B-A22B-Instruct-2507",
919
- "hf:Qwen/Qwen3-235B-A22B-Thinking-2507",
920
- "hf:Qwen/Qwen3-Coder-480B-A35B-Instruct",
921
- "hf:deepseek-ai/DeepSeek-R1",
922
- "hf:deepseek-ai/DeepSeek-R1-0528",
923
- "hf:deepseek-ai/DeepSeek-V3",
924
- "hf:deepseek-ai/DeepSeek-V3-0324",
925
- "hf:deepseek-ai/DeepSeek-V3.1",
926
- "hf:deepseek-ai/DeepSeek-V3.1-Terminus",
927
- "hf:meta-llama/Llama-3.1-405B-Instruct",
928
- "hf:meta-llama/Llama-3.1-70B-Instruct",
929
- "hf:meta-llama/Llama-3.1-8B-Instruct",
930
- "hf:meta-llama/Llama-3.3-70B-Instruct",
931
- "hf:meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8",
932
- "hf:meta-llama/Llama-4-Scout-17B-16E-Instruct",
933
- "hf:moonshotai/Kimi-K2-Instruct",
934
- "hf:moonshotai/Kimi-K2-Instruct-0905",
935
- "hf:openai/gpt-oss-120b",
936
- "hf:zai-org/GLM-4.5",
937
- "hf:zai-org/GLM-4.6"
701
+ "xai/grok-3",
702
+ "xai/grok-3-mini"
938
703
  ],
939
- docUrl: "https://synthetic.new/pricing",
704
+ docUrl: "https://docs.github.com/en/github-models",
940
705
  gateway: "models.dev"
941
706
  },
942
707
  deepinfra: {
@@ -953,92 +718,92 @@ var provider_registry_default = {
953
718
  docUrl: "https://deepinfra.com/models",
954
719
  gateway: "models.dev"
955
720
  },
956
- zhipuai: {
957
- url: "https://open.bigmodel.cn/api/paas/v4",
721
+ "fireworks-ai": {
722
+ url: "https://api.fireworks.ai/inference/v1/",
723
+ apiKeyEnvVar: "FIREWORKS_API_KEY",
724
+ apiKeyHeader: "Authorization",
725
+ name: "Fireworks AI",
726
+ models: [
727
+ "accounts/fireworks/models/deepseek-r1-0528",
728
+ "accounts/fireworks/models/deepseek-v3-0324",
729
+ "accounts/fireworks/models/deepseek-v3p1",
730
+ "accounts/fireworks/models/glm-4p5",
731
+ "accounts/fireworks/models/glm-4p5-air",
732
+ "accounts/fireworks/models/gpt-oss-120b",
733
+ "accounts/fireworks/models/gpt-oss-20b",
734
+ "accounts/fireworks/models/kimi-k2-instruct",
735
+ "accounts/fireworks/models/qwen3-235b-a22b",
736
+ "accounts/fireworks/models/qwen3-coder-480b-a35b-instruct"
737
+ ],
738
+ docUrl: "https://fireworks.ai/docs/",
739
+ gateway: "models.dev"
740
+ },
741
+ zai: {
742
+ url: "https://api.z.ai/api/paas/v4",
958
743
  apiKeyEnvVar: "ZHIPU_API_KEY",
959
744
  apiKeyHeader: "Authorization",
960
- name: "Zhipu AI",
745
+ name: "Z.AI",
961
746
  models: ["glm-4.5", "glm-4.5-air", "glm-4.5-flash", "glm-4.5v", "glm-4.6"],
962
747
  docUrl: "https://docs.z.ai/guides/overview/pricing",
963
748
  gateway: "models.dev"
964
749
  },
965
- submodel: {
966
- url: "https://llm.submodel.ai/v1",
967
- apiKeyEnvVar: "SUBMODEL_INSTAGEN_ACCESS_KEY",
750
+ opencode: {
751
+ url: "https://opencode.ai/zen/v1",
752
+ apiKeyEnvVar: "OPENCODE_API_KEY",
968
753
  apiKeyHeader: "Authorization",
969
- name: "submodel",
754
+ name: "OpenCode Zen",
970
755
  models: [
971
- "Qwen/Qwen3-235B-A22B-Instruct-2507",
972
- "Qwen/Qwen3-235B-A22B-Thinking-2507",
973
- "Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8",
974
- "deepseek-ai/DeepSeek-R1-0528",
975
- "deepseek-ai/DeepSeek-V3-0324",
976
- "deepseek-ai/DeepSeek-V3.1",
977
- "openai/gpt-oss-120b",
978
- "zai-org/GLM-4.5-Air",
979
- "zai-org/GLM-4.5-FP8"
756
+ "an-gbt",
757
+ "big-pickle",
758
+ "claude-3-5-haiku",
759
+ "claude-haiku-4-5",
760
+ "claude-opus-4-1",
761
+ "claude-sonnet-4",
762
+ "claude-sonnet-4-5",
763
+ "code-supernova",
764
+ "glm-4.6",
765
+ "gpt-5",
766
+ "gpt-5-codex",
767
+ "grok-code",
768
+ "kimi-k2",
769
+ "qwen3-coder"
980
770
  ],
981
- docUrl: "https://submodel.gitbook.io",
771
+ docUrl: "https://opencode.ai/docs/zen",
982
772
  gateway: "models.dev"
983
773
  },
984
- zai: {
985
- url: "https://api.z.ai/api/paas/v4",
774
+ "zhipuai-coding-plan": {
775
+ url: "https://open.bigmodel.cn/api/coding/paas/v4",
986
776
  apiKeyEnvVar: "ZHIPU_API_KEY",
987
777
  apiKeyHeader: "Authorization",
988
- name: "Z.AI",
778
+ name: "Zhipu AI Coding Plan",
989
779
  models: ["glm-4.5", "glm-4.5-air", "glm-4.5-flash", "glm-4.5v", "glm-4.6"],
990
- docUrl: "https://docs.z.ai/guides/overview/pricing",
780
+ docUrl: "https://docs.bigmodel.cn/cn/coding-plan/overview",
991
781
  gateway: "models.dev"
992
782
  },
993
- inference: {
994
- url: "https://inference.net/v1",
995
- apiKeyEnvVar: "INFERENCE_API_KEY",
783
+ inception: {
784
+ url: "https://api.inceptionlabs.ai/v1/",
785
+ apiKeyEnvVar: "INCEPTION_API_KEY",
996
786
  apiKeyHeader: "Authorization",
997
- name: "Inference",
998
- models: [
999
- "google/gemma-3",
1000
- "meta/llama-3.1-8b-instruct",
1001
- "meta/llama-3.2-11b-vision-instruct",
1002
- "meta/llama-3.2-1b-instruct",
1003
- "meta/llama-3.2-3b-instruct",
1004
- "mistral/mistral-nemo-12b-instruct",
1005
- "osmosis/osmosis-structure-0.6b",
1006
- "qwen/qwen-2.5-7b-vision-instruct",
1007
- "qwen/qwen3-embedding-4b"
1008
- ],
1009
- docUrl: "https://inference.net/models",
787
+ name: "Inception",
788
+ models: ["mercury", "mercury-coder"],
789
+ docUrl: "https://platform.inceptionlabs.ai/docs",
1010
790
  gateway: "models.dev"
1011
791
  },
1012
- requesty: {
1013
- url: "https://router.requesty.ai/v1",
1014
- apiKeyEnvVar: "REQUESTY_API_KEY",
792
+ llama: {
793
+ url: "https://api.llama.com/compat/v1/",
794
+ apiKeyEnvVar: "LLAMA_API_KEY",
1015
795
  apiKeyHeader: "Authorization",
1016
- name: "Requesty",
796
+ name: "Llama",
1017
797
  models: [
1018
- "anthropic/claude-3-7-sonnet",
1019
- "anthropic/claude-4-sonnet-20250522",
1020
- "anthropic/claude-opus-4",
1021
- "anthropic/claude-opus-4-1-20250805",
1022
- "google/gemini-2.5-flash",
1023
- "google/gemini-2.5-pro",
1024
- "openai/gpt-4.1",
1025
- "openai/gpt-4.1-mini",
1026
- "openai/gpt-4o-mini",
1027
- "openai/gpt-5",
1028
- "openai/gpt-5-mini",
1029
- "openai/gpt-5-nano",
1030
- "openai/o4-mini"
798
+ "cerebras-llama-4-maverick-17b-128e-instruct",
799
+ "cerebras-llama-4-scout-17b-16e-instruct",
800
+ "groq-llama-4-maverick-17b-128e-instruct",
801
+ "llama-3.3-70b-instruct",
802
+ "llama-3.3-8b-instruct",
803
+ "llama-4-maverick-17b-128e-instruct-fp8",
804
+ "llama-4-scout-17b-16e-instruct-fp8"
1031
805
  ],
1032
- docUrl: "https://requesty.ai/solution/llm-routing/models",
1033
- gateway: "models.dev"
1034
- },
1035
- morph: {
1036
- url: "https://api.morphllm.com/v1",
1037
- apiKeyEnvVar: "MORPH_API_KEY",
1038
- apiKeyHeader: "Authorization",
1039
- name: "Morph",
1040
- models: ["auto", "morph-v3-fast", "morph-v3-large"],
1041
- docUrl: "https://docs.morphllm.com/api-reference/introduction",
806
+ docUrl: "https://llama.developer.meta.com/docs/models",
1042
807
  gateway: "models.dev"
1043
808
  },
1044
809
  lmstudio: {
@@ -1050,97 +815,359 @@ var provider_registry_default = {
1050
815
  docUrl: "https://lmstudio.ai/models",
1051
816
  gateway: "models.dev"
1052
817
  },
818
+ openai: {
819
+ apiKeyEnvVar: "OPENAI_API_KEY",
820
+ name: "OpenAI",
821
+ models: [
822
+ "codex-mini-latest",
823
+ "gpt-3.5-turbo",
824
+ "gpt-4",
825
+ "gpt-4-turbo",
826
+ "gpt-4.1",
827
+ "gpt-4.1-mini",
828
+ "gpt-4.1-nano",
829
+ "gpt-4o",
830
+ "gpt-4o-2024-05-13",
831
+ "gpt-4o-2024-08-06",
832
+ "gpt-4o-2024-11-20",
833
+ "gpt-4o-mini",
834
+ "gpt-5",
835
+ "gpt-5-chat-latest",
836
+ "gpt-5-codex",
837
+ "gpt-5-mini",
838
+ "gpt-5-nano",
839
+ "o1",
840
+ "o1-mini",
841
+ "o1-preview",
842
+ "o1-pro",
843
+ "o3",
844
+ "o3-deep-research",
845
+ "o3-mini",
846
+ "o3-pro",
847
+ "o4-mini",
848
+ "o4-mini-deep-research"
849
+ ],
850
+ docUrl: "https://platform.openai.com/docs/models",
851
+ gateway: "models.dev"
852
+ },
1053
853
  anthropic: {
1054
854
  apiKeyEnvVar: "ANTHROPIC_API_KEY",
1055
855
  name: "Anthropic",
1056
856
  models: [
1057
857
  "claude-3-5-haiku-20241022",
858
+ "claude-3-5-haiku-latest",
1058
859
  "claude-3-5-sonnet-20240620",
1059
860
  "claude-3-5-sonnet-20241022",
1060
861
  "claude-3-7-sonnet-20250219",
862
+ "claude-3-7-sonnet-latest",
1061
863
  "claude-3-haiku-20240307",
1062
864
  "claude-3-opus-20240229",
1063
865
  "claude-3-sonnet-20240229",
866
+ "claude-haiku-4-5",
1064
867
  "claude-haiku-4-5-20251001",
868
+ "claude-opus-4-0",
869
+ "claude-opus-4-1",
1065
870
  "claude-opus-4-1-20250805",
1066
871
  "claude-opus-4-20250514",
872
+ "claude-sonnet-4-0",
1067
873
  "claude-sonnet-4-20250514",
874
+ "claude-sonnet-4-5",
1068
875
  "claude-sonnet-4-5-20250929"
1069
876
  ],
1070
877
  docUrl: "https://docs.anthropic.com/en/docs/about-claude/models",
1071
878
  gateway: "models.dev"
1072
879
  },
1073
- "fireworks-ai": {
1074
- url: "https://api.fireworks.ai/inference/v1/",
1075
- apiKeyEnvVar: "FIREWORKS_API_KEY",
880
+ moonshotai: {
881
+ url: "https://api.moonshot.ai/v1",
882
+ apiKeyEnvVar: "MOONSHOT_API_KEY",
1076
883
  apiKeyHeader: "Authorization",
1077
- name: "Fireworks AI",
1078
- models: [
1079
- "accounts/fireworks/models/deepseek-r1-0528",
1080
- "accounts/fireworks/models/deepseek-v3-0324",
1081
- "accounts/fireworks/models/deepseek-v3p1",
1082
- "accounts/fireworks/models/glm-4p5",
1083
- "accounts/fireworks/models/glm-4p5-air",
1084
- "accounts/fireworks/models/gpt-oss-120b",
1085
- "accounts/fireworks/models/gpt-oss-20b",
1086
- "accounts/fireworks/models/kimi-k2-instruct",
1087
- "accounts/fireworks/models/qwen3-235b-a22b",
1088
- "accounts/fireworks/models/qwen3-coder-480b-a35b-instruct"
1089
- ],
1090
- docUrl: "https://fireworks.ai/docs/",
884
+ name: "Moonshot AI",
885
+ models: ["kimi-k2-0711-preview", "kimi-k2-0905-preview", "kimi-k2-turbo-preview"],
886
+ docUrl: "https://platform.moonshot.ai/docs/api/chat",
1091
887
  gateway: "models.dev"
1092
888
  },
1093
- modelscope: {
1094
- url: "https://api-inference.modelscope.cn/v1",
1095
- apiKeyEnvVar: "MODELSCOPE_API_KEY",
889
+ inference: {
890
+ url: "https://inference.net/v1",
891
+ apiKeyEnvVar: "INFERENCE_API_KEY",
1096
892
  apiKeyHeader: "Authorization",
1097
- name: "ModelScope",
893
+ name: "Inference",
1098
894
  models: [
1099
- "Qwen/Qwen3-235B-A22B-Instruct-2507",
1100
- "Qwen/Qwen3-235B-A22B-Thinking-2507",
1101
- "Qwen/Qwen3-30B-A3B-Instruct-2507",
1102
- "Qwen/Qwen3-30B-A3B-Thinking-2507",
1103
- "Qwen/Qwen3-Coder-30B-A3B-Instruct",
1104
- "ZhipuAI/GLM-4.5"
895
+ "google/gemma-3",
896
+ "meta/llama-3.1-8b-instruct",
897
+ "meta/llama-3.2-11b-vision-instruct",
898
+ "meta/llama-3.2-1b-instruct",
899
+ "meta/llama-3.2-3b-instruct",
900
+ "mistral/mistral-nemo-12b-instruct",
901
+ "osmosis/osmosis-structure-0.6b",
902
+ "qwen/qwen-2.5-7b-vision-instruct",
903
+ "qwen/qwen3-embedding-4b"
1105
904
  ],
1106
- docUrl: "https://modelscope.cn/docs/model-service/API-Inference/intro",
905
+ docUrl: "https://inference.net/models",
1107
906
  gateway: "models.dev"
1108
907
  },
1109
- llama: {
1110
- url: "https://api.llama.com/compat/v1/",
1111
- apiKeyEnvVar: "LLAMA_API_KEY",
908
+ baseten: {
909
+ url: "https://inference.baseten.co/v1",
910
+ apiKeyEnvVar: "BASETEN_API_KEY",
1112
911
  apiKeyHeader: "Authorization",
1113
- name: "Llama",
912
+ name: "Baseten",
913
+ models: ["Qwen3/Qwen3-Coder-480B-A35B-Instruct", "moonshotai/Kimi-K2-Instruct-0905", "zai-org/GLM-4.6"],
914
+ docUrl: "https://docs.baseten.co/development/model-apis/overview",
915
+ gateway: "models.dev"
916
+ },
917
+ cortecs: {
918
+ url: "https://api.cortecs.ai/v1",
919
+ apiKeyEnvVar: "CORTECS_API_KEY",
920
+ apiKeyHeader: "Authorization",
921
+ name: "Cortecs",
1114
922
  models: [
1115
- "cerebras-llama-4-maverick-17b-128e-instruct",
1116
- "cerebras-llama-4-scout-17b-16e-instruct",
1117
- "groq-llama-4-maverick-17b-128e-instruct",
1118
- "llama-3.3-70b-instruct",
1119
- "llama-3.3-8b-instruct",
1120
- "llama-4-maverick-17b-128e-instruct-fp8",
1121
- "llama-4-scout-17b-16e-instruct-fp8"
923
+ "claude-4-5-sonnet",
924
+ "claude-sonnet-4",
925
+ "deepseek-v3-0324",
926
+ "gemini-2.5-pro",
927
+ "gpt-4.1",
928
+ "gpt-oss-120b",
929
+ "kimi-k2-instruct",
930
+ "llama-3.1-405b-instruct",
931
+ "nova-pro-v1",
932
+ "qwen3-32b",
933
+ "qwen3-coder-480b-a35b-instruct"
1122
934
  ],
1123
- docUrl: "https://llama.developer.meta.com/docs/models",
935
+ docUrl: "https://api.cortecs.ai/v1/models",
1124
936
  gateway: "models.dev"
1125
937
  },
1126
- cerebras: {
1127
- url: "https://api.cerebras.ai/v1",
1128
- apiKeyEnvVar: "CEREBRAS_API_KEY",
938
+ mistral: {
939
+ url: "https://api.mistral.ai/v1",
940
+ apiKeyEnvVar: "MISTRAL_API_KEY",
1129
941
  apiKeyHeader: "Authorization",
1130
- name: "Cerebras",
1131
- models: ["gpt-oss-120b", "qwen-3-235b-a22b-instruct-2507", "qwen-3-coder-480b"],
1132
- docUrl: "https://inference-docs.cerebras.ai/models/overview",
942
+ name: "Mistral",
943
+ models: [
944
+ "codestral-latest",
945
+ "devstral-medium-2507",
946
+ "devstral-small-2505",
947
+ "devstral-small-2507",
948
+ "magistral-medium-latest",
949
+ "magistral-small",
950
+ "ministral-3b-latest",
951
+ "ministral-8b-latest",
952
+ "mistral-large-latest",
953
+ "mistral-medium-2505",
954
+ "mistral-medium-2508",
955
+ "mistral-medium-latest",
956
+ "mistral-nemo",
957
+ "mistral-small-latest",
958
+ "open-mistral-7b",
959
+ "open-mixtral-8x22b",
960
+ "open-mixtral-8x7b",
961
+ "pixtral-12b",
962
+ "pixtral-large-latest"
963
+ ],
964
+ docUrl: "https://docs.mistral.ai/getting-started/models/",
1133
965
  gateway: "models.dev"
1134
966
  },
1135
- netlify: {
1136
- apiKeyEnvVar: ["NETLIFY_TOKEN", "NETLIFY_SITE_ID"],
967
+ venice: {
968
+ url: "https://api.venice.ai/api/v1",
969
+ apiKeyEnvVar: "VENICE_API_KEY",
1137
970
  apiKeyHeader: "Authorization",
1138
- name: "Netlify",
1139
- gateway: "netlify",
971
+ name: "Venice AI",
1140
972
  models: [
1141
- "anthropic/claude-3-5-haiku-20241022",
1142
- "anthropic/claude-3-5-haiku-latest",
1143
- "anthropic/claude-3-7-sonnet-20250219",
973
+ "deepseek-coder-v2-lite",
974
+ "deepseek-r1-671b",
975
+ "dolphin-2.9.2-qwen2-72b",
976
+ "llama-3.1-405b",
977
+ "llama-3.2-3b",
978
+ "llama-3.3-70b",
979
+ "mistral-31-24b",
980
+ "qwen-2.5-coder-32b",
981
+ "qwen-2.5-qwq-32b",
982
+ "qwen-2.5-vl",
983
+ "qwen3-235b",
984
+ "qwen3-4b",
985
+ "venice-uncensored"
986
+ ],
987
+ docUrl: "https://docs.venice.ai",
988
+ gateway: "models.dev"
989
+ },
990
+ alibaba: {
991
+ url: "https://dashscope-intl.aliyuncs.com/compatible-mode/v1",
992
+ apiKeyEnvVar: "DASHSCOPE_API_KEY",
993
+ apiKeyHeader: "Authorization",
994
+ name: "Alibaba",
995
+ models: [
996
+ "qvq-max",
997
+ "qwen-flash",
998
+ "qwen-max",
999
+ "qwen-mt-plus",
1000
+ "qwen-mt-turbo",
1001
+ "qwen-omni-turbo",
1002
+ "qwen-omni-turbo-realtime",
1003
+ "qwen-plus",
1004
+ "qwen-plus-character-ja",
1005
+ "qwen-turbo",
1006
+ "qwen-vl-max",
1007
+ "qwen-vl-ocr",
1008
+ "qwen-vl-plus",
1009
+ "qwen2-5-14b-instruct",
1010
+ "qwen2-5-32b-instruct",
1011
+ "qwen2-5-72b-instruct",
1012
+ "qwen2-5-7b-instruct",
1013
+ "qwen2-5-omni-7b",
1014
+ "qwen2-5-vl-72b-instruct",
1015
+ "qwen2-5-vl-7b-instruct",
1016
+ "qwen3-14b",
1017
+ "qwen3-235b-a22b",
1018
+ "qwen3-32b",
1019
+ "qwen3-8b",
1020
+ "qwen3-asr-flash",
1021
+ "qwen3-coder-30b-a3b-instruct",
1022
+ "qwen3-coder-480b-a35b-instruct",
1023
+ "qwen3-coder-flash",
1024
+ "qwen3-coder-plus",
1025
+ "qwen3-livetranslate-flash-realtime",
1026
+ "qwen3-max",
1027
+ "qwen3-next-80b-a3b-instruct",
1028
+ "qwen3-next-80b-a3b-thinking",
1029
+ "qwen3-omni-flash",
1030
+ "qwen3-omni-flash-realtime",
1031
+ "qwen3-vl-235b-a22b",
1032
+ "qwen3-vl-30b-a3b",
1033
+ "qwen3-vl-plus",
1034
+ "qwq-plus"
1035
+ ],
1036
+ docUrl: "https://www.alibabacloud.com/help/en/model-studio/models",
1037
+ gateway: "models.dev"
1038
+ },
1039
+ "alibaba-cn": {
1040
+ url: "https://dashscope.aliyuncs.com/compatible-mode/v1",
1041
+ apiKeyEnvVar: "DASHSCOPE_API_KEY",
1042
+ apiKeyHeader: "Authorization",
1043
+ name: "Alibaba (China)",
1044
+ models: [
1045
+ "deepseek-r1",
1046
+ "deepseek-r1-0528",
1047
+ "deepseek-r1-distill-llama-70b",
1048
+ "deepseek-r1-distill-llama-8b",
1049
+ "deepseek-r1-distill-qwen-1-5b",
1050
+ "deepseek-r1-distill-qwen-14b",
1051
+ "deepseek-r1-distill-qwen-32b",
1052
+ "deepseek-r1-distill-qwen-7b",
1053
+ "deepseek-v3",
1054
+ "deepseek-v3-1",
1055
+ "deepseek-v3-2-exp",
1056
+ "moonshot-kimi-k2-instruct",
1057
+ "qvq-max",
1058
+ "qwen-deep-research",
1059
+ "qwen-doc-turbo",
1060
+ "qwen-flash",
1061
+ "qwen-long",
1062
+ "qwen-math-plus",
1063
+ "qwen-math-turbo",
1064
+ "qwen-max",
1065
+ "qwen-mt-plus",
1066
+ "qwen-mt-turbo",
1067
+ "qwen-omni-turbo",
1068
+ "qwen-omni-turbo-realtime",
1069
+ "qwen-plus",
1070
+ "qwen-plus-character",
1071
+ "qwen-turbo",
1072
+ "qwen-vl-max",
1073
+ "qwen-vl-ocr",
1074
+ "qwen-vl-plus",
1075
+ "qwen2-5-14b-instruct",
1076
+ "qwen2-5-32b-instruct",
1077
+ "qwen2-5-72b-instruct",
1078
+ "qwen2-5-7b-instruct",
1079
+ "qwen2-5-coder-32b-instruct",
1080
+ "qwen2-5-coder-7b-instruct",
1081
+ "qwen2-5-math-72b-instruct",
1082
+ "qwen2-5-math-7b-instruct",
1083
+ "qwen2-5-omni-7b",
1084
+ "qwen2-5-vl-72b-instruct",
1085
+ "qwen2-5-vl-7b-instruct",
1086
+ "qwen3-14b",
1087
+ "qwen3-235b-a22b",
1088
+ "qwen3-32b",
1089
+ "qwen3-8b",
1090
+ "qwen3-asr-flash",
1091
+ "qwen3-coder-30b-a3b-instruct",
1092
+ "qwen3-coder-480b-a35b-instruct",
1093
+ "qwen3-coder-flash",
1094
+ "qwen3-coder-plus",
1095
+ "qwen3-max",
1096
+ "qwen3-next-80b-a3b-instruct",
1097
+ "qwen3-next-80b-a3b-thinking",
1098
+ "qwen3-omni-flash",
1099
+ "qwen3-omni-flash-realtime",
1100
+ "qwen3-vl-235b-a22b",
1101
+ "qwen3-vl-30b-a3b",
1102
+ "qwen3-vl-plus",
1103
+ "qwq-32b",
1104
+ "qwq-plus",
1105
+ "tongyi-intent-detect-v3"
1106
+ ],
1107
+ docUrl: "https://www.alibabacloud.com/help/en/model-studio/models",
1108
+ gateway: "models.dev"
1109
+ },
1110
+ chutes: {
1111
+ url: "https://llm.chutes.ai/v1",
1112
+ apiKeyEnvVar: "CHUTES_API_KEY",
1113
+ apiKeyHeader: "Authorization",
1114
+ name: "Chutes",
1115
+ models: [
1116
+ "Qwen/Qwen3-235B-A22B-Instruct-2507",
1117
+ "Qwen/Qwen3-235B-A22B-Thinking-2507",
1118
+ "Qwen/Qwen3-30B-A3B",
1119
+ "Qwen/Qwen3-30B-A3B-Instruct-2507",
1120
+ "Qwen/Qwen3-30B-A3B-Thinking-2507",
1121
+ "Qwen/Qwen3-Coder-30B-A3B-Instruct",
1122
+ "Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8",
1123
+ "Qwen/Qwen3-Next-80B-A3B-Instruct",
1124
+ "Qwen/Qwen3-Next-80B-A3B-Thinking",
1125
+ "chutesai/Devstral-Small-2505",
1126
+ "chutesai/Mistral-Small-3.2-24B-Instruct-2506",
1127
+ "deepseek-ai/DeepSeek-R1-0528",
1128
+ "deepseek-ai/DeepSeek-R1-0528-Qwen3-8B",
1129
+ "deepseek-ai/DeepSeek-R1-Distill-Llama-70B",
1130
+ "deepseek-ai/DeepSeek-V3-0324",
1131
+ "deepseek-ai/DeepSeek-V3.1",
1132
+ "deepseek-ai/DeepSeek-V3.1-Terminus",
1133
+ "deepseek-ai/DeepSeek-V3.1-turbo",
1134
+ "deepseek-ai/DeepSeek-V3.1:THINKING",
1135
+ "deepseek-ai/DeepSeek-V3.2-Exp",
1136
+ "meituan-longcat/LongCat-Flash-Chat-FP8",
1137
+ "moonshotai/Kimi-Dev-72B",
1138
+ "moonshotai/Kimi-K2-Instruct-0905",
1139
+ "moonshotai/Kimi-K2-Instruct-75k",
1140
+ "moonshotai/Kimi-VL-A3B-Thinking",
1141
+ "openai/gpt-oss-120b",
1142
+ "tngtech/DeepSeek-R1T-Chimera",
1143
+ "tngtech/DeepSeek-TNG-R1T2-Chimera",
1144
+ "zai-org/GLM-4.5-Air",
1145
+ "zai-org/GLM-4.5-FP8",
1146
+ "zai-org/GLM-4.5-turbo",
1147
+ "zai-org/GLM-4.6-FP8",
1148
+ "zai-org/GLM-4.6-turbo"
1149
+ ],
1150
+ docUrl: "https://llm.chutes.ai/v1/models",
1151
+ gateway: "models.dev"
1152
+ },
1153
+ upstage: {
1154
+ url: "https://api.upstage.ai",
1155
+ apiKeyEnvVar: "UPSTAGE_API_KEY",
1156
+ apiKeyHeader: "Authorization",
1157
+ name: "Upstage",
1158
+ models: ["solar-mini", "solar-pro2"],
1159
+ docUrl: "https://developers.upstage.ai/docs/apis/chat",
1160
+ gateway: "models.dev"
1161
+ },
1162
+ netlify: {
1163
+ apiKeyEnvVar: ["NETLIFY_TOKEN", "NETLIFY_SITE_ID"],
1164
+ apiKeyHeader: "Authorization",
1165
+ name: "Netlify",
1166
+ gateway: "netlify",
1167
+ models: [
1168
+ "anthropic/claude-3-5-haiku-20241022",
1169
+ "anthropic/claude-3-5-haiku-latest",
1170
+ "anthropic/claude-3-7-sonnet-20250219",
1144
1171
  "anthropic/claude-3-7-sonnet-latest",
1145
1172
  "anthropic/claude-3-haiku-20240307",
1146
1173
  "anthropic/claude-haiku-4-5-20251001",
@@ -1177,51 +1204,58 @@ var provider_registry_default = {
1177
1204
  }
1178
1205
  },
1179
1206
  models: {
1180
- "moonshotai-cn": ["kimi-k2-0711-preview", "kimi-k2-0905-preview", "kimi-k2-turbo-preview"],
1207
+ nvidia: [
1208
+ "black-forest-labs/flux.1-dev",
1209
+ "deepseek-ai/deepseek-v3.1",
1210
+ "deepseek-ai/deepseek-v3.1-terminus",
1211
+ "google/gemma-3-27b-it",
1212
+ "microsoft/phi-4-mini-instruct",
1213
+ "moonshotai/kimi-k2-instruct",
1214
+ "moonshotai/kimi-k2-instruct-0905",
1215
+ "nvidia/cosmos-nemotron-34b",
1216
+ "nvidia/llama-3.1-nemotron-ultra-253b-v1",
1217
+ "nvidia/nemoretriever-ocr-v1",
1218
+ "nvidia/parakeet-tdt-0.6b-v2",
1219
+ "openai/gpt-oss-120b",
1220
+ "openai/whisper-large-v3",
1221
+ "qwen/qwen3-235b-a22b",
1222
+ "qwen/qwen3-coder-480b-a35b-instruct"
1223
+ ],
1224
+ modelscope: [
1225
+ "Qwen/Qwen3-235B-A22B-Instruct-2507",
1226
+ "Qwen/Qwen3-235B-A22B-Thinking-2507",
1227
+ "Qwen/Qwen3-30B-A3B-Instruct-2507",
1228
+ "Qwen/Qwen3-30B-A3B-Thinking-2507",
1229
+ "Qwen/Qwen3-Coder-30B-A3B-Instruct",
1230
+ "ZhipuAI/GLM-4.5",
1231
+ "ZhipuAI/GLM-4.6"
1232
+ ],
1233
+ aihubmix: [
1234
+ "DeepSeek-V3.2-Exp",
1235
+ "DeepSeek-V3.2-Exp-Think",
1236
+ "Kimi-K2-0905",
1237
+ "claude-haiku-4-5",
1238
+ "claude-opus-4-1",
1239
+ "claude-sonnet-4-5",
1240
+ "gemini-2.5-flash",
1241
+ "gemini-2.5-pro",
1242
+ "glm-4.6",
1243
+ "gpt-4.1",
1244
+ "gpt-4.1-mini",
1245
+ "gpt-4.1-nano",
1246
+ "gpt-4o",
1247
+ "gpt-4o-2024-11-20",
1248
+ "gpt-5",
1249
+ "gpt-5-codex",
1250
+ "gpt-5-mini",
1251
+ "gpt-5-nano",
1252
+ "gpt-5-pro",
1253
+ "o4-mini",
1254
+ "qwen3-235b-a22b-instruct-2507",
1255
+ "qwen3-235b-a22b-thinking-2507",
1256
+ "qwen3-coder-480b-a35b-instruct"
1257
+ ],
1181
1258
  lucidquery: ["lucidnova-rf1-100b", "lucidquery-nexus-coder"],
1182
- moonshotai: ["kimi-k2-0711-preview", "kimi-k2-0905-preview", "kimi-k2-turbo-preview"],
1183
- "zai-coding-plan": ["glm-4.5", "glm-4.5-air", "glm-4.5-flash", "glm-4.5v", "glm-4.6"],
1184
- alibaba: [
1185
- "qvq-max",
1186
- "qwen-flash",
1187
- "qwen-max",
1188
- "qwen-mt-plus",
1189
- "qwen-mt-turbo",
1190
- "qwen-omni-turbo",
1191
- "qwen-omni-turbo-realtime",
1192
- "qwen-plus",
1193
- "qwen-plus-character-ja",
1194
- "qwen-turbo",
1195
- "qwen-vl-max",
1196
- "qwen-vl-ocr",
1197
- "qwen-vl-plus",
1198
- "qwen2-5-14b-instruct",
1199
- "qwen2-5-32b-instruct",
1200
- "qwen2-5-72b-instruct",
1201
- "qwen2-5-7b-instruct",
1202
- "qwen2-5-omni-7b",
1203
- "qwen2-5-vl-72b-instruct",
1204
- "qwen2-5-vl-7b-instruct",
1205
- "qwen3-14b",
1206
- "qwen3-235b-a22b",
1207
- "qwen3-32b",
1208
- "qwen3-8b",
1209
- "qwen3-asr-flash",
1210
- "qwen3-coder-30b-a3b-instruct",
1211
- "qwen3-coder-480b-a35b-instruct",
1212
- "qwen3-coder-flash",
1213
- "qwen3-coder-plus",
1214
- "qwen3-livetranslate-flash-realtime",
1215
- "qwen3-max",
1216
- "qwen3-next-80b-a3b-instruct",
1217
- "qwen3-next-80b-a3b-thinking",
1218
- "qwen3-omni-flash",
1219
- "qwen3-omni-flash-realtime",
1220
- "qwen3-vl-235b-a22b",
1221
- "qwen3-vl-30b-a3b",
1222
- "qwen3-vl-plus",
1223
- "qwq-plus"
1224
- ],
1225
1259
  xai: [
1226
1260
  "grok-2",
1227
1261
  "grok-2-1212",
@@ -1244,293 +1278,370 @@ var provider_registry_default = {
1244
1278
  "grok-code-fast-1",
1245
1279
  "grok-vision-beta"
1246
1280
  ],
1247
- nvidia: [
1248
- "black-forest-labs/flux.1-dev",
1249
- "deepseek-ai/deepseek-v3.1",
1250
- "google/gemma-3-27b-it",
1251
- "microsoft/phi-4-mini-instruct",
1281
+ deepseek: ["deepseek-chat", "deepseek-reasoner"],
1282
+ requesty: [
1283
+ "anthropic/claude-3-7-sonnet",
1284
+ "anthropic/claude-4-sonnet-20250522",
1285
+ "anthropic/claude-opus-4",
1286
+ "anthropic/claude-opus-4-1-20250805",
1287
+ "google/gemini-2.5-flash",
1288
+ "google/gemini-2.5-pro",
1289
+ "openai/gpt-4.1",
1290
+ "openai/gpt-4.1-mini",
1291
+ "openai/gpt-4o-mini",
1292
+ "openai/gpt-5",
1293
+ "openai/gpt-5-mini",
1294
+ "openai/gpt-5-nano",
1295
+ "openai/o4-mini"
1296
+ ],
1297
+ nebius: [
1298
+ "NousResearch/hermes-4-405b",
1299
+ "NousResearch/hermes-4-70b",
1300
+ "deepseek-ai/deepseek-v3",
1301
+ "meta-llama/llama-3.3-70b-instruct-base",
1302
+ "meta-llama/llama-3.3-70b-instruct-fast",
1303
+ "meta-llama/llama-3_1-405b-instruct",
1252
1304
  "moonshotai/kimi-k2-instruct",
1253
- "moonshotai/kimi-k2-instruct-0905",
1254
- "nvidia/cosmos-nemotron-34b",
1255
- "nvidia/llama-3.1-nemotron-ultra-253b-v1",
1256
- "nvidia/nemoretriever-ocr-v1",
1257
- "nvidia/parakeet-tdt-0.6b-v2",
1305
+ "nvidia/llama-3_1-nemotron-ultra-253b-v1",
1258
1306
  "openai/gpt-oss-120b",
1259
- "openai/whisper-large-v3",
1260
- "qwen/qwen3-235b-a22b",
1261
- "qwen/qwen3-coder-480b-a35b-instruct"
1307
+ "openai/gpt-oss-20b",
1308
+ "qwen/qwen3-235b-a22b-instruct-2507",
1309
+ "qwen/qwen3-235b-a22b-thinking-2507",
1310
+ "qwen/qwen3-coder-480b-a35b-instruct",
1311
+ "zai-org/glm-4.5",
1312
+ "zai-org/glm-4.5-air"
1262
1313
  ],
1263
- upstage: ["solar-mini", "solar-pro2"],
1264
- groq: [
1265
- "deepseek-r1-distill-llama-70b",
1266
- "gemma2-9b-it",
1267
- "llama-3.1-8b-instant",
1268
- "llama-3.3-70b-versatile",
1269
- "llama-guard-3-8b",
1270
- "llama3-70b-8192",
1271
- "llama3-8b-8192",
1272
- "meta-llama/llama-4-maverick-17b-128e-instruct",
1273
- "meta-llama/llama-4-scout-17b-16e-instruct",
1274
- "meta-llama/llama-guard-4-12b",
1275
- "mistral-saba-24b",
1276
- "moonshotai/kimi-k2-instruct",
1277
- "moonshotai/kimi-k2-instruct-0905",
1314
+ fastrouter: [
1315
+ "anthropic/claude-opus-4.1",
1316
+ "anthropic/claude-sonnet-4",
1317
+ "deepseek-ai/deepseek-r1-distill-llama-70b",
1318
+ "google/gemini-2.5-flash",
1319
+ "google/gemini-2.5-pro",
1320
+ "moonshotai/kimi-k2",
1321
+ "openai/gpt-4.1",
1322
+ "openai/gpt-5",
1323
+ "openai/gpt-5-mini",
1324
+ "openai/gpt-5-nano",
1278
1325
  "openai/gpt-oss-120b",
1279
1326
  "openai/gpt-oss-20b",
1280
- "qwen-qwq-32b",
1281
- "qwen/qwen3-32b"
1327
+ "qwen/qwen3-coder",
1328
+ "x-ai/grok-4"
1282
1329
  ],
1283
- mistral: [
1284
- "codestral-latest",
1285
- "devstral-medium-2507",
1286
- "devstral-small-2505",
1287
- "devstral-small-2507",
1288
- "magistral-medium-latest",
1289
- "magistral-small",
1290
- "ministral-3b-latest",
1291
- "ministral-8b-latest",
1292
- "mistral-large-latest",
1293
- "mistral-medium-2505",
1294
- "mistral-medium-2508",
1295
- "mistral-medium-latest",
1296
- "mistral-nemo",
1297
- "mistral-small-latest",
1298
- "open-mistral-7b",
1299
- "open-mixtral-8x22b",
1300
- "open-mixtral-8x7b",
1301
- "pixtral-12b",
1302
- "pixtral-large-latest"
1330
+ perplexity: ["sonar", "sonar-pro", "sonar-reasoning", "sonar-reasoning-pro"],
1331
+ morph: ["auto", "morph-v3-fast", "morph-v3-large"],
1332
+ "zai-coding-plan": ["glm-4.5", "glm-4.5-air", "glm-4.5-flash", "glm-4.5v", "glm-4.6"],
1333
+ huggingface: [
1334
+ "Qwen/Qwen3-235B-A22B-Thinking-2507",
1335
+ "Qwen/Qwen3-Coder-480B-A35B-Instruct",
1336
+ "Qwen/Qwen3-Next-80B-A3B-Instruct",
1337
+ "Qwen/Qwen3-Next-80B-A3B-Thinking",
1338
+ "deepseek-ai/DeepSeek-R1-0528",
1339
+ "deepseek-ai/Deepseek-V3-0324",
1340
+ "moonshotai/Kimi-K2-Instruct",
1341
+ "moonshotai/Kimi-K2-Instruct-0905",
1342
+ "zai-org/GLM-4.5",
1343
+ "zai-org/GLM-4.5-Air",
1344
+ "zai-org/GLM-4.6"
1303
1345
  ],
1304
- vercel: [
1305
- "alibaba/qwen3-coder-plus",
1306
- "alibaba/qwen3-max",
1307
- "alibaba/qwen3-next-80b-a3b-instruct",
1308
- "alibaba/qwen3-next-80b-a3b-thinking",
1309
- "alibaba/qwen3-vl-instruct",
1310
- "alibaba/qwen3-vl-thinking",
1311
- "amazon/nova-lite",
1312
- "amazon/nova-micro",
1313
- "amazon/nova-pro",
1314
- "anthropic/claude-3-5-haiku",
1315
- "anthropic/claude-3-haiku",
1316
- "anthropic/claude-3-opus",
1317
- "anthropic/claude-3.5-sonnet",
1346
+ openrouter: [
1347
+ "anthropic/claude-3.5-haiku",
1318
1348
  "anthropic/claude-3.7-sonnet",
1319
- "anthropic/claude-4-1-opus",
1320
- "anthropic/claude-4-opus",
1321
- "anthropic/claude-4-sonnet",
1322
- "anthropic/claude-4.5-sonnet",
1323
1349
  "anthropic/claude-haiku-4.5",
1324
- "cerebras/qwen3-coder",
1325
- "deepseek/deepseek-r1",
1350
+ "anthropic/claude-opus-4",
1351
+ "anthropic/claude-opus-4.1",
1352
+ "anthropic/claude-sonnet-4",
1353
+ "anthropic/claude-sonnet-4.5",
1354
+ "cognitivecomputations/dolphin3.0-mistral-24b",
1355
+ "cognitivecomputations/dolphin3.0-r1-mistral-24b",
1356
+ "deepseek/deepseek-chat-v3-0324",
1357
+ "deepseek/deepseek-chat-v3.1",
1358
+ "deepseek/deepseek-r1-0528-qwen3-8b:free",
1359
+ "deepseek/deepseek-r1-0528:free",
1326
1360
  "deepseek/deepseek-r1-distill-llama-70b",
1361
+ "deepseek/deepseek-r1-distill-qwen-14b",
1362
+ "deepseek/deepseek-r1:free",
1363
+ "deepseek/deepseek-v3-base:free",
1327
1364
  "deepseek/deepseek-v3.1-terminus",
1328
- "deepseek/deepseek-v3.2-exp",
1329
- "deepseek/deepseek-v3.2-exp-thinking",
1330
- "google/gemini-2.0-flash",
1331
- "google/gemini-2.0-flash-lite",
1365
+ "featherless/qwerky-72b",
1366
+ "google/gemini-2.0-flash-001",
1367
+ "google/gemini-2.0-flash-exp:free",
1332
1368
  "google/gemini-2.5-flash",
1333
1369
  "google/gemini-2.5-flash-lite",
1334
1370
  "google/gemini-2.5-flash-lite-preview-09-2025",
1335
1371
  "google/gemini-2.5-flash-preview-09-2025",
1336
1372
  "google/gemini-2.5-pro",
1337
- "meta/llama-3.3-70b",
1338
- "meta/llama-4-maverick",
1339
- "meta/llama-4-scout",
1340
- "mistral/codestral",
1341
- "mistral/magistral-medium",
1342
- "mistral/magistral-small",
1343
- "mistral/ministral-3b",
1344
- "mistral/ministral-8b",
1345
- "mistral/mistral-large",
1346
- "mistral/mistral-small",
1347
- "mistral/mixtral-8x22b-instruct",
1348
- "mistral/pixtral-12b",
1349
- "mistral/pixtral-large",
1373
+ "google/gemini-2.5-pro-preview-05-06",
1374
+ "google/gemini-2.5-pro-preview-06-05",
1375
+ "google/gemma-2-9b-it:free",
1376
+ "google/gemma-3-12b-it",
1377
+ "google/gemma-3-27b-it",
1378
+ "google/gemma-3n-e4b-it",
1379
+ "google/gemma-3n-e4b-it:free",
1380
+ "meta-llama/llama-3.2-11b-vision-instruct",
1381
+ "meta-llama/llama-3.3-70b-instruct:free",
1382
+ "meta-llama/llama-4-scout:free",
1383
+ "microsoft/mai-ds-r1:free",
1384
+ "mistralai/codestral-2508",
1385
+ "mistralai/devstral-medium-2507",
1386
+ "mistralai/devstral-small-2505",
1387
+ "mistralai/devstral-small-2505:free",
1388
+ "mistralai/devstral-small-2507",
1389
+ "mistralai/mistral-7b-instruct:free",
1390
+ "mistralai/mistral-medium-3",
1391
+ "mistralai/mistral-medium-3.1",
1392
+ "mistralai/mistral-nemo:free",
1393
+ "mistralai/mistral-small-3.1-24b-instruct",
1394
+ "mistralai/mistral-small-3.2-24b-instruct",
1395
+ "mistralai/mistral-small-3.2-24b-instruct:free",
1396
+ "moonshotai/kimi-dev-72b:free",
1350
1397
  "moonshotai/kimi-k2",
1351
- "morph/morph-v3-fast",
1352
- "morph/morph-v3-large",
1353
- "openai/gpt-4-turbo",
1398
+ "moonshotai/kimi-k2-0905",
1399
+ "moonshotai/kimi-k2:free",
1400
+ "nousresearch/deephermes-3-llama-3-8b-preview",
1401
+ "nousresearch/hermes-4-405b",
1402
+ "nousresearch/hermes-4-70b",
1354
1403
  "openai/gpt-4.1",
1355
1404
  "openai/gpt-4.1-mini",
1356
- "openai/gpt-4.1-nano",
1357
- "openai/gpt-4o",
1358
1405
  "openai/gpt-4o-mini",
1359
1406
  "openai/gpt-5",
1407
+ "openai/gpt-5-chat",
1360
1408
  "openai/gpt-5-codex",
1409
+ "openai/gpt-5-image",
1361
1410
  "openai/gpt-5-mini",
1362
1411
  "openai/gpt-5-nano",
1363
1412
  "openai/gpt-oss-120b",
1364
1413
  "openai/gpt-oss-20b",
1365
- "openai/o1",
1366
- "openai/o3",
1367
- "openai/o3-mini",
1368
1414
  "openai/o4-mini",
1369
- "perplexity/sonar",
1370
- "perplexity/sonar-pro",
1371
- "perplexity/sonar-reasoning",
1372
- "perplexity/sonar-reasoning-pro",
1373
- "vercel/v0-1.0-md",
1374
- "vercel/v0-1.5-md",
1375
- "xai/grok-2",
1376
- "xai/grok-2-vision",
1377
- "xai/grok-3",
1378
- "xai/grok-3-fast",
1379
- "xai/grok-3-mini",
1380
- "xai/grok-3-mini-fast",
1381
- "xai/grok-4",
1382
- "xai/grok-4-fast",
1383
- "xai/grok-4-fast-non-reasoning",
1384
- "xai/grok-code-fast-1",
1385
- "zai/glm-4.5",
1386
- "zai/glm-4.5-air",
1387
- "zai/glm-4.5v",
1388
- "zai/glm-4.6"
1389
- ],
1390
- nebius: [
1391
- "NousResearch/hermes-4-405b",
1392
- "NousResearch/hermes-4-70b",
1393
- "deepseek-ai/deepseek-v3",
1394
- "meta-llama/llama-3.3-70b-instruct-base",
1395
- "meta-llama/llama-3.3-70b-instruct-fast",
1396
- "meta-llama/llama-3_1-405b-instruct",
1397
- "moonshotai/kimi-k2-instruct",
1398
- "nvidia/llama-3_1-nemotron-ultra-253b-v1",
1399
- "openai/gpt-oss-120b",
1400
- "openai/gpt-oss-20b",
1401
- "qwen/qwen3-235b-a22b-instruct-2507",
1415
+ "openrouter/cypher-alpha:free",
1416
+ "openrouter/horizon-alpha",
1417
+ "openrouter/horizon-beta",
1418
+ "openrouter/sonoma-dusk-alpha",
1419
+ "openrouter/sonoma-sky-alpha",
1420
+ "qwen/qwen-2.5-coder-32b-instruct",
1421
+ "qwen/qwen2.5-vl-32b-instruct:free",
1422
+ "qwen/qwen2.5-vl-72b-instruct",
1423
+ "qwen/qwen2.5-vl-72b-instruct:free",
1424
+ "qwen/qwen3-14b:free",
1425
+ "qwen/qwen3-235b-a22b-07-25",
1426
+ "qwen/qwen3-235b-a22b-07-25:free",
1402
1427
  "qwen/qwen3-235b-a22b-thinking-2507",
1403
- "qwen/qwen3-coder-480b-a35b-instruct",
1404
- "zai-org/glm-4.5",
1405
- "zai-org/glm-4.5-air"
1406
- ],
1407
- deepseek: ["deepseek-chat", "deepseek-reasoner"],
1408
- "alibaba-cn": [
1409
- "deepseek-r1",
1410
- "deepseek-r1-0528",
1411
- "deepseek-r1-distill-llama-70b",
1412
- "deepseek-r1-distill-llama-8b",
1413
- "deepseek-r1-distill-qwen-1-5b",
1414
- "deepseek-r1-distill-qwen-14b",
1415
- "deepseek-r1-distill-qwen-32b",
1416
- "deepseek-r1-distill-qwen-7b",
1417
- "deepseek-v3",
1418
- "deepseek-v3-1",
1419
- "deepseek-v3-2-exp",
1420
- "moonshot-kimi-k2-instruct",
1421
- "qvq-max",
1422
- "qwen-deep-research",
1423
- "qwen-doc-turbo",
1424
- "qwen-flash",
1425
- "qwen-long",
1426
- "qwen-math-plus",
1427
- "qwen-math-turbo",
1428
- "qwen-max",
1429
- "qwen-mt-plus",
1430
- "qwen-mt-turbo",
1431
- "qwen-omni-turbo",
1432
- "qwen-omni-turbo-realtime",
1433
- "qwen-plus",
1434
- "qwen-plus-character",
1435
- "qwen-turbo",
1436
- "qwen-vl-max",
1437
- "qwen-vl-ocr",
1438
- "qwen-vl-plus",
1439
- "qwen2-5-14b-instruct",
1440
- "qwen2-5-32b-instruct",
1441
- "qwen2-5-72b-instruct",
1442
- "qwen2-5-7b-instruct",
1443
- "qwen2-5-coder-32b-instruct",
1444
- "qwen2-5-coder-7b-instruct",
1445
- "qwen2-5-math-72b-instruct",
1446
- "qwen2-5-math-7b-instruct",
1447
- "qwen2-5-omni-7b",
1448
- "qwen2-5-vl-72b-instruct",
1449
- "qwen2-5-vl-7b-instruct",
1450
- "qwen3-14b",
1451
- "qwen3-235b-a22b",
1452
- "qwen3-32b",
1453
- "qwen3-8b",
1454
- "qwen3-asr-flash",
1455
- "qwen3-coder-30b-a3b-instruct",
1456
- "qwen3-coder-480b-a35b-instruct",
1457
- "qwen3-coder-flash",
1458
- "qwen3-coder-plus",
1459
- "qwen3-max",
1460
- "qwen3-next-80b-a3b-instruct",
1461
- "qwen3-next-80b-a3b-thinking",
1462
- "qwen3-omni-flash",
1463
- "qwen3-omni-flash-realtime",
1464
- "qwen3-vl-235b-a22b",
1465
- "qwen3-vl-30b-a3b",
1466
- "qwen3-vl-plus",
1467
- "qwq-32b",
1468
- "qwq-plus",
1469
- "tongyi-intent-detect-v3"
1428
+ "qwen/qwen3-235b-a22b:free",
1429
+ "qwen/qwen3-30b-a3b-instruct-2507",
1430
+ "qwen/qwen3-30b-a3b-thinking-2507",
1431
+ "qwen/qwen3-30b-a3b:free",
1432
+ "qwen/qwen3-32b:free",
1433
+ "qwen/qwen3-8b:free",
1434
+ "qwen/qwen3-coder",
1435
+ "qwen/qwen3-coder:free",
1436
+ "qwen/qwen3-max",
1437
+ "qwen/qwen3-next-80b-a3b-instruct",
1438
+ "qwen/qwen3-next-80b-a3b-thinking",
1439
+ "qwen/qwq-32b:free",
1440
+ "rekaai/reka-flash-3",
1441
+ "sarvamai/sarvam-m:free",
1442
+ "thudm/glm-z1-32b:free",
1443
+ "tngtech/deepseek-r1t2-chimera:free",
1444
+ "x-ai/grok-3",
1445
+ "x-ai/grok-3-beta",
1446
+ "x-ai/grok-3-mini",
1447
+ "x-ai/grok-3-mini-beta",
1448
+ "x-ai/grok-4",
1449
+ "x-ai/grok-4-fast",
1450
+ "x-ai/grok-4-fast:free",
1451
+ "x-ai/grok-code-fast-1",
1452
+ "z-ai/glm-4.5",
1453
+ "z-ai/glm-4.5-air",
1454
+ "z-ai/glm-4.5-air:free",
1455
+ "z-ai/glm-4.5v",
1456
+ "z-ai/glm-4.6"
1470
1457
  ],
1471
- venice: [
1472
- "deepseek-coder-v2-lite",
1473
- "deepseek-r1-671b",
1474
- "dolphin-2.9.2-qwen2-72b",
1475
- "llama-3.1-405b",
1476
- "llama-3.2-3b",
1477
- "llama-3.3-70b",
1478
- "mistral-31-24b",
1479
- "qwen-2.5-coder-32b",
1480
- "qwen-2.5-qwq-32b",
1481
- "qwen-2.5-vl",
1482
- "qwen3-235b",
1483
- "qwen3-4b",
1484
- "venice-uncensored"
1458
+ cerebras: ["gpt-oss-120b", "qwen-3-235b-a22b-instruct-2507", "qwen-3-coder-480b"],
1459
+ zhipuai: ["glm-4.5", "glm-4.5-air", "glm-4.5-flash", "glm-4.5v", "glm-4.6"],
1460
+ togetherai: [
1461
+ "Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8",
1462
+ "deepseek-ai/DeepSeek-R1",
1463
+ "deepseek-ai/DeepSeek-V3",
1464
+ "meta-llama/Llama-3.3-70B-Instruct-Turbo",
1465
+ "moonshotai/Kimi-K2-Instruct",
1466
+ "openai/gpt-oss-120b"
1485
1467
  ],
1486
- chutes: [
1468
+ wandb: [
1487
1469
  "Qwen/Qwen3-235B-A22B-Instruct-2507",
1488
1470
  "Qwen/Qwen3-235B-A22B-Thinking-2507",
1489
- "Qwen/Qwen3-30B-A3B",
1490
- "Qwen/Qwen3-30B-A3B-Instruct-2507",
1491
- "Qwen/Qwen3-30B-A3B-Thinking-2507",
1492
- "Qwen/Qwen3-Coder-30B-A3B-Instruct",
1493
- "Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8",
1494
- "Qwen/Qwen3-Next-80B-A3B-Instruct",
1495
- "Qwen/Qwen3-Next-80B-A3B-Thinking",
1496
- "chutesai/Devstral-Small-2505",
1497
- "chutesai/Mistral-Small-3.2-24B-Instruct-2506",
1471
+ "Qwen/Qwen3-Coder-480B-A35B-Instruct",
1498
1472
  "deepseek-ai/DeepSeek-R1-0528",
1499
- "deepseek-ai/DeepSeek-R1-0528-Qwen3-8B",
1500
- "deepseek-ai/DeepSeek-R1-Distill-Llama-70B",
1501
1473
  "deepseek-ai/DeepSeek-V3-0324",
1502
- "deepseek-ai/DeepSeek-V3.1",
1503
- "deepseek-ai/DeepSeek-V3.1-Terminus",
1504
- "deepseek-ai/DeepSeek-V3.1-turbo",
1505
- "deepseek-ai/DeepSeek-V3.1:THINKING",
1506
- "deepseek-ai/DeepSeek-V3.2-Exp",
1507
- "meituan-longcat/LongCat-Flash-Chat-FP8",
1508
- "moonshotai/Kimi-Dev-72B",
1509
- "moonshotai/Kimi-K2-Instruct-0905",
1510
- "moonshotai/Kimi-K2-Instruct-75k",
1511
- "moonshotai/Kimi-VL-A3B-Thinking",
1474
+ "meta-llama/Llama-3.1-8B-Instruct",
1475
+ "meta-llama/Llama-3.3-70B-Instruct",
1476
+ "meta-llama/Llama-4-Scout-17B-16E-Instruct",
1477
+ "microsoft/Phi-4-mini-instruct",
1478
+ "moonshotai/Kimi-K2-Instruct"
1479
+ ],
1480
+ submodel: [
1481
+ "Qwen/Qwen3-235B-A22B-Instruct-2507",
1482
+ "Qwen/Qwen3-235B-A22B-Thinking-2507",
1483
+ "Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8",
1484
+ "deepseek-ai/DeepSeek-R1-0528",
1485
+ "deepseek-ai/DeepSeek-V3-0324",
1486
+ "deepseek-ai/DeepSeek-V3.1",
1512
1487
  "openai/gpt-oss-120b",
1513
- "tngtech/DeepSeek-R1T-Chimera",
1514
- "tngtech/DeepSeek-TNG-R1T2-Chimera",
1515
1488
  "zai-org/GLM-4.5-Air",
1516
- "zai-org/GLM-4.5-FP8",
1517
- "zai-org/GLM-4.5-turbo",
1518
- "zai-org/GLM-4.6-FP8",
1519
- "zai-org/GLM-4.6-turbo"
1489
+ "zai-org/GLM-4.5-FP8"
1520
1490
  ],
1521
- cortecs: [
1522
- "claude-4-5-sonnet",
1523
- "claude-sonnet-4",
1524
- "deepseek-v3-0324",
1491
+ vercel: [
1492
+ "alibaba/qwen3-coder-plus",
1493
+ "alibaba/qwen3-max",
1494
+ "alibaba/qwen3-next-80b-a3b-instruct",
1495
+ "alibaba/qwen3-next-80b-a3b-thinking",
1496
+ "alibaba/qwen3-vl-instruct",
1497
+ "alibaba/qwen3-vl-thinking",
1498
+ "amazon/nova-lite",
1499
+ "amazon/nova-micro",
1500
+ "amazon/nova-pro",
1501
+ "anthropic/claude-3-5-haiku",
1502
+ "anthropic/claude-3-haiku",
1503
+ "anthropic/claude-3-opus",
1504
+ "anthropic/claude-3.5-sonnet",
1505
+ "anthropic/claude-3.7-sonnet",
1506
+ "anthropic/claude-4-1-opus",
1507
+ "anthropic/claude-4-opus",
1508
+ "anthropic/claude-4-sonnet",
1509
+ "anthropic/claude-4.5-sonnet",
1510
+ "anthropic/claude-haiku-4.5",
1511
+ "cerebras/qwen3-coder",
1512
+ "deepseek/deepseek-r1",
1513
+ "deepseek/deepseek-r1-distill-llama-70b",
1514
+ "deepseek/deepseek-v3.1-terminus",
1515
+ "deepseek/deepseek-v3.2-exp",
1516
+ "deepseek/deepseek-v3.2-exp-thinking",
1517
+ "google/gemini-2.0-flash",
1518
+ "google/gemini-2.0-flash-lite",
1519
+ "google/gemini-2.5-flash",
1520
+ "google/gemini-2.5-flash-lite",
1521
+ "google/gemini-2.5-flash-lite-preview-09-2025",
1522
+ "google/gemini-2.5-flash-preview-09-2025",
1523
+ "google/gemini-2.5-pro",
1524
+ "meta/llama-3.3-70b",
1525
+ "meta/llama-4-maverick",
1526
+ "meta/llama-4-scout",
1527
+ "mistral/codestral",
1528
+ "mistral/magistral-medium",
1529
+ "mistral/magistral-small",
1530
+ "mistral/ministral-3b",
1531
+ "mistral/ministral-8b",
1532
+ "mistral/mistral-large",
1533
+ "mistral/mistral-small",
1534
+ "mistral/mixtral-8x22b-instruct",
1535
+ "mistral/pixtral-12b",
1536
+ "mistral/pixtral-large",
1537
+ "moonshotai/kimi-k2",
1538
+ "morph/morph-v3-fast",
1539
+ "morph/morph-v3-large",
1540
+ "openai/gpt-4-turbo",
1541
+ "openai/gpt-4.1",
1542
+ "openai/gpt-4.1-mini",
1543
+ "openai/gpt-4.1-nano",
1544
+ "openai/gpt-4o",
1545
+ "openai/gpt-4o-mini",
1546
+ "openai/gpt-5",
1547
+ "openai/gpt-5-codex",
1548
+ "openai/gpt-5-mini",
1549
+ "openai/gpt-5-nano",
1550
+ "openai/gpt-oss-120b",
1551
+ "openai/gpt-oss-20b",
1552
+ "openai/o1",
1553
+ "openai/o3",
1554
+ "openai/o3-mini",
1555
+ "openai/o4-mini",
1556
+ "perplexity/sonar",
1557
+ "perplexity/sonar-pro",
1558
+ "perplexity/sonar-reasoning",
1559
+ "perplexity/sonar-reasoning-pro",
1560
+ "vercel/v0-1.0-md",
1561
+ "vercel/v0-1.5-md",
1562
+ "xai/grok-2",
1563
+ "xai/grok-2-vision",
1564
+ "xai/grok-3",
1565
+ "xai/grok-3-fast",
1566
+ "xai/grok-3-mini",
1567
+ "xai/grok-3-mini-fast",
1568
+ "xai/grok-4",
1569
+ "xai/grok-4-fast",
1570
+ "xai/grok-4-fast-non-reasoning",
1571
+ "xai/grok-code-fast-1",
1572
+ "zai/glm-4.5",
1573
+ "zai/glm-4.5-air",
1574
+ "zai/glm-4.5v",
1575
+ "zai/glm-4.6"
1576
+ ],
1577
+ synthetic: [
1578
+ "hf:Qwen/Qwen2.5-Coder-32B-Instruct",
1579
+ "hf:Qwen/Qwen3-235B-A22B-Instruct-2507",
1580
+ "hf:Qwen/Qwen3-235B-A22B-Thinking-2507",
1581
+ "hf:Qwen/Qwen3-Coder-480B-A35B-Instruct",
1582
+ "hf:deepseek-ai/DeepSeek-R1",
1583
+ "hf:deepseek-ai/DeepSeek-R1-0528",
1584
+ "hf:deepseek-ai/DeepSeek-V3",
1585
+ "hf:deepseek-ai/DeepSeek-V3-0324",
1586
+ "hf:deepseek-ai/DeepSeek-V3.1",
1587
+ "hf:deepseek-ai/DeepSeek-V3.1-Terminus",
1588
+ "hf:meta-llama/Llama-3.1-405B-Instruct",
1589
+ "hf:meta-llama/Llama-3.1-70B-Instruct",
1590
+ "hf:meta-llama/Llama-3.1-8B-Instruct",
1591
+ "hf:meta-llama/Llama-3.3-70B-Instruct",
1592
+ "hf:meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8",
1593
+ "hf:meta-llama/Llama-4-Scout-17B-16E-Instruct",
1594
+ "hf:moonshotai/Kimi-K2-Instruct",
1595
+ "hf:moonshotai/Kimi-K2-Instruct-0905",
1596
+ "hf:openai/gpt-oss-120b",
1597
+ "hf:zai-org/GLM-4.5",
1598
+ "hf:zai-org/GLM-4.6"
1599
+ ],
1600
+ google: [
1601
+ "gemini-1.5-flash",
1602
+ "gemini-1.5-flash-8b",
1603
+ "gemini-1.5-pro",
1604
+ "gemini-2.0-flash",
1605
+ "gemini-2.0-flash-lite",
1606
+ "gemini-2.5-flash",
1607
+ "gemini-2.5-flash-image",
1608
+ "gemini-2.5-flash-image-preview",
1609
+ "gemini-2.5-flash-lite",
1610
+ "gemini-2.5-flash-lite-preview-06-17",
1611
+ "gemini-2.5-flash-lite-preview-09-2025",
1612
+ "gemini-2.5-flash-preview-04-17",
1613
+ "gemini-2.5-flash-preview-05-20",
1614
+ "gemini-2.5-flash-preview-09-2025",
1615
+ "gemini-2.5-flash-preview-tts",
1525
1616
  "gemini-2.5-pro",
1526
- "gpt-4.1",
1527
- "gpt-oss-120b",
1528
- "kimi-k2-instruct",
1529
- "llama-3.1-405b-instruct",
1530
- "nova-pro-v1",
1531
- "qwen3-32b",
1532
- "qwen3-coder-480b-a35b-instruct"
1617
+ "gemini-2.5-pro-preview-05-06",
1618
+ "gemini-2.5-pro-preview-06-05",
1619
+ "gemini-2.5-pro-preview-tts",
1620
+ "gemini-flash-latest",
1621
+ "gemini-flash-lite-latest",
1622
+ "gemini-live-2.5-flash",
1623
+ "gemini-live-2.5-flash-preview-native-audio"
1624
+ ],
1625
+ groq: [
1626
+ "deepseek-r1-distill-llama-70b",
1627
+ "gemma2-9b-it",
1628
+ "llama-3.1-8b-instant",
1629
+ "llama-3.3-70b-versatile",
1630
+ "llama-guard-3-8b",
1631
+ "llama3-70b-8192",
1632
+ "llama3-8b-8192",
1633
+ "meta-llama/llama-4-maverick-17b-128e-instruct",
1634
+ "meta-llama/llama-4-scout-17b-16e-instruct",
1635
+ "meta-llama/llama-guard-4-12b",
1636
+ "mistral-saba-24b",
1637
+ "moonshotai/kimi-k2-instruct",
1638
+ "moonshotai/kimi-k2-instruct-0905",
1639
+ "openai/gpt-oss-120b",
1640
+ "openai/gpt-oss-20b",
1641
+ "qwen-qwq-32b",
1642
+ "qwen/qwen3-32b"
1533
1643
  ],
1644
+ "moonshotai-cn": ["kimi-k2-0711-preview", "kimi-k2-0905-preview", "kimi-k2-turbo-preview"],
1534
1645
  "github-models": [
1535
1646
  "ai21-labs/ai21-jamba-1.5-large",
1536
1647
  "ai21-labs/ai21-jamba-1.5-mini",
@@ -1588,97 +1699,53 @@ var provider_registry_default = {
1588
1699
  "xai/grok-3",
1589
1700
  "xai/grok-3-mini"
1590
1701
  ],
1591
- togetherai: [
1592
- "Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8",
1593
- "deepseek-ai/DeepSeek-R1",
1594
- "deepseek-ai/DeepSeek-V3",
1595
- "meta-llama/Llama-3.3-70B-Instruct-Turbo",
1596
- "moonshotai/Kimi-K2-Instruct",
1597
- "openai/gpt-oss-120b"
1598
- ],
1599
- baseten: ["Qwen3/Qwen3-Coder-480B-A35B-Instruct", "moonshotai/Kimi-K2-Instruct-0905"],
1600
- huggingface: [
1601
- "Qwen/Qwen3-235B-A22B-Thinking-2507",
1702
+ deepinfra: [
1602
1703
  "Qwen/Qwen3-Coder-480B-A35B-Instruct",
1603
- "Qwen/Qwen3-Next-80B-A3B-Instruct",
1604
- "Qwen/Qwen3-Next-80B-A3B-Thinking",
1605
- "deepseek-ai/DeepSeek-R1-0528",
1606
- "deepseek-ai/Deepseek-V3-0324",
1704
+ "Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo",
1607
1705
  "moonshotai/Kimi-K2-Instruct",
1608
- "moonshotai/Kimi-K2-Instruct-0905",
1609
- "zai-org/GLM-4.5",
1610
- "zai-org/GLM-4.5-Air",
1611
- "zai-org/GLM-4.6"
1612
- ],
1613
- opencode: [
1614
- "an-g8x",
1615
- "claude-3-5-haiku",
1616
- "claude-haiku-4-5",
1617
- "claude-opus-4-1",
1618
- "claude-sonnet-4",
1619
- "claude-sonnet-4-5",
1620
- "code-supernova",
1621
- "glm-4.6",
1622
- "gpt-5",
1623
- "gpt-5-codex",
1624
- "grok-code",
1625
- "kimi-k2",
1626
- "qwen3-coder"
1627
- ],
1628
- fastrouter: [
1629
- "anthropic/claude-opus-4.1",
1630
- "anthropic/claude-sonnet-4",
1631
- "deepseek-ai/deepseek-r1-distill-llama-70b",
1632
- "google/gemini-2.5-flash",
1633
- "google/gemini-2.5-pro",
1634
- "moonshotai/kimi-k2",
1635
- "openai/gpt-4.1",
1636
- "openai/gpt-5",
1637
- "openai/gpt-5-mini",
1638
- "openai/gpt-5-nano",
1639
- "openai/gpt-oss-120b",
1640
- "openai/gpt-oss-20b",
1641
- "qwen/qwen3-coder",
1642
- "x-ai/grok-4"
1706
+ "zai-org/GLM-4.5"
1643
1707
  ],
1644
- google: [
1645
- "gemini-1.5-flash",
1646
- "gemini-1.5-flash-8b",
1647
- "gemini-1.5-pro",
1648
- "gemini-2.0-flash",
1649
- "gemini-2.0-flash-lite",
1650
- "gemini-2.5-flash",
1651
- "gemini-2.5-flash-image",
1652
- "gemini-2.5-flash-image-preview",
1653
- "gemini-2.5-flash-lite",
1654
- "gemini-2.5-flash-lite-preview-06-17",
1655
- "gemini-2.5-flash-lite-preview-09-2025",
1656
- "gemini-2.5-flash-preview-04-17",
1657
- "gemini-2.5-flash-preview-05-20",
1658
- "gemini-2.5-flash-preview-09-2025",
1659
- "gemini-2.5-flash-preview-tts",
1660
- "gemini-2.5-pro",
1661
- "gemini-2.5-pro-preview-05-06",
1662
- "gemini-2.5-pro-preview-06-05",
1663
- "gemini-2.5-pro-preview-tts",
1664
- "gemini-flash-latest",
1665
- "gemini-flash-lite-latest",
1666
- "gemini-live-2.5-flash",
1667
- "gemini-live-2.5-flash-preview-native-audio"
1708
+ "fireworks-ai": [
1709
+ "accounts/fireworks/models/deepseek-r1-0528",
1710
+ "accounts/fireworks/models/deepseek-v3-0324",
1711
+ "accounts/fireworks/models/deepseek-v3p1",
1712
+ "accounts/fireworks/models/glm-4p5",
1713
+ "accounts/fireworks/models/glm-4p5-air",
1714
+ "accounts/fireworks/models/gpt-oss-120b",
1715
+ "accounts/fireworks/models/gpt-oss-20b",
1716
+ "accounts/fireworks/models/kimi-k2-instruct",
1717
+ "accounts/fireworks/models/qwen3-235b-a22b",
1718
+ "accounts/fireworks/models/qwen3-coder-480b-a35b-instruct"
1719
+ ],
1720
+ zai: ["glm-4.5", "glm-4.5-air", "glm-4.5-flash", "glm-4.5v", "glm-4.6"],
1721
+ opencode: [
1722
+ "an-gbt",
1723
+ "big-pickle",
1724
+ "claude-3-5-haiku",
1725
+ "claude-haiku-4-5",
1726
+ "claude-opus-4-1",
1727
+ "claude-sonnet-4",
1728
+ "claude-sonnet-4-5",
1729
+ "code-supernova",
1730
+ "glm-4.6",
1731
+ "gpt-5",
1732
+ "gpt-5-codex",
1733
+ "grok-code",
1734
+ "kimi-k2",
1735
+ "qwen3-coder"
1668
1736
  ],
1737
+ "zhipuai-coding-plan": ["glm-4.5", "glm-4.5-air", "glm-4.5-flash", "glm-4.5v", "glm-4.6"],
1669
1738
  inception: ["mercury", "mercury-coder"],
1670
- wandb: [
1671
- "Qwen/Qwen3-235B-A22B-Instruct-2507",
1672
- "Qwen/Qwen3-235B-A22B-Thinking-2507",
1673
- "Qwen/Qwen3-Coder-480B-A35B-Instruct",
1674
- "deepseek-ai/DeepSeek-R1-0528",
1675
- "deepseek-ai/DeepSeek-V3-0324",
1676
- "meta-llama/Llama-3.1-8B-Instruct",
1677
- "meta-llama/Llama-3.3-70B-Instruct",
1678
- "meta-llama/Llama-4-Scout-17B-16E-Instruct",
1679
- "microsoft/Phi-4-mini-instruct",
1680
- "moonshotai/Kimi-K2-Instruct"
1739
+ llama: [
1740
+ "cerebras-llama-4-maverick-17b-128e-instruct",
1741
+ "cerebras-llama-4-scout-17b-16e-instruct",
1742
+ "groq-llama-4-maverick-17b-128e-instruct",
1743
+ "llama-3.3-70b-instruct",
1744
+ "llama-3.3-8b-instruct",
1745
+ "llama-4-maverick-17b-128e-instruct-fp8",
1746
+ "llama-4-scout-17b-16e-instruct-fp8"
1681
1747
  ],
1748
+ lmstudio: ["openai/gpt-oss-20b", "qwen/qwen3-30b-a3b-2507", "qwen/qwen3-coder-30b"],
1682
1749
  openai: [
1683
1750
  "codex-mini-latest",
1684
1751
  "gpt-3.5-turbo",
@@ -1708,234 +1775,229 @@ var provider_registry_default = {
1708
1775
  "o4-mini",
1709
1776
  "o4-mini-deep-research"
1710
1777
  ],
1711
- "zhipuai-coding-plan": ["glm-4.5", "glm-4.5-air", "glm-4.5-flash", "glm-4.5v", "glm-4.6"],
1712
- perplexity: ["sonar", "sonar-pro", "sonar-reasoning", "sonar-reasoning-pro"],
1713
- openrouter: [
1714
- "anthropic/claude-3.5-haiku",
1715
- "anthropic/claude-3.7-sonnet",
1716
- "anthropic/claude-4.5-haiku",
1717
- "anthropic/claude-opus-4",
1718
- "anthropic/claude-opus-4.1",
1719
- "anthropic/claude-sonnet-4",
1720
- "anthropic/claude-sonnet-4.5",
1721
- "cognitivecomputations/dolphin3.0-mistral-24b",
1722
- "cognitivecomputations/dolphin3.0-r1-mistral-24b",
1723
- "deepseek/deepseek-chat-v3-0324",
1724
- "deepseek/deepseek-chat-v3.1",
1725
- "deepseek/deepseek-r1-0528-qwen3-8b:free",
1726
- "deepseek/deepseek-r1-0528:free",
1727
- "deepseek/deepseek-r1-distill-llama-70b",
1728
- "deepseek/deepseek-r1-distill-qwen-14b",
1729
- "deepseek/deepseek-r1:free",
1730
- "deepseek/deepseek-v3-base:free",
1731
- "deepseek/deepseek-v3.1-terminus",
1732
- "featherless/qwerky-72b",
1733
- "google/gemini-2.0-flash-001",
1734
- "google/gemini-2.0-flash-exp:free",
1735
- "google/gemini-2.5-flash",
1736
- "google/gemini-2.5-flash-lite",
1737
- "google/gemini-2.5-flash-lite-preview-09-2025",
1738
- "google/gemini-2.5-flash-preview-09-2025",
1739
- "google/gemini-2.5-pro",
1740
- "google/gemini-2.5-pro-preview-05-06",
1741
- "google/gemini-2.5-pro-preview-06-05",
1742
- "google/gemma-2-9b-it:free",
1743
- "google/gemma-3-12b-it",
1744
- "google/gemma-3-27b-it",
1745
- "google/gemma-3n-e4b-it",
1746
- "google/gemma-3n-e4b-it:free",
1747
- "meta-llama/llama-3.2-11b-vision-instruct",
1748
- "meta-llama/llama-3.3-70b-instruct:free",
1749
- "meta-llama/llama-4-scout:free",
1750
- "microsoft/mai-ds-r1:free",
1751
- "mistralai/codestral-2508",
1752
- "mistralai/devstral-medium-2507",
1753
- "mistralai/devstral-small-2505",
1754
- "mistralai/devstral-small-2505:free",
1755
- "mistralai/devstral-small-2507",
1756
- "mistralai/mistral-7b-instruct:free",
1757
- "mistralai/mistral-medium-3",
1758
- "mistralai/mistral-medium-3.1",
1759
- "mistralai/mistral-nemo:free",
1760
- "mistralai/mistral-small-3.1-24b-instruct",
1761
- "mistralai/mistral-small-3.2-24b-instruct",
1762
- "mistralai/mistral-small-3.2-24b-instruct:free",
1763
- "moonshotai/kimi-dev-72b:free",
1764
- "moonshotai/kimi-k2",
1765
- "moonshotai/kimi-k2-0905",
1766
- "moonshotai/kimi-k2:free",
1767
- "nousresearch/deephermes-3-llama-3-8b-preview",
1768
- "nousresearch/hermes-4-405b",
1769
- "nousresearch/hermes-4-70b",
1770
- "openai/gpt-4.1",
1771
- "openai/gpt-4.1-mini",
1772
- "openai/gpt-4o-mini",
1773
- "openai/gpt-5",
1774
- "openai/gpt-5-chat",
1775
- "openai/gpt-5-codex",
1776
- "openai/gpt-5-image",
1777
- "openai/gpt-5-mini",
1778
- "openai/gpt-5-nano",
1779
- "openai/gpt-oss-120b",
1780
- "openai/gpt-oss-20b",
1781
- "openai/o4-mini",
1782
- "openrouter/cypher-alpha:free",
1783
- "openrouter/horizon-alpha",
1784
- "openrouter/horizon-beta",
1785
- "openrouter/sonoma-dusk-alpha",
1786
- "openrouter/sonoma-sky-alpha",
1787
- "qwen/qwen-2.5-coder-32b-instruct",
1788
- "qwen/qwen2.5-vl-32b-instruct:free",
1789
- "qwen/qwen2.5-vl-72b-instruct",
1790
- "qwen/qwen2.5-vl-72b-instruct:free",
1791
- "qwen/qwen3-14b:free",
1792
- "qwen/qwen3-235b-a22b-07-25",
1793
- "qwen/qwen3-235b-a22b-07-25:free",
1794
- "qwen/qwen3-235b-a22b-thinking-2507",
1795
- "qwen/qwen3-235b-a22b:free",
1796
- "qwen/qwen3-30b-a3b-instruct-2507",
1797
- "qwen/qwen3-30b-a3b-thinking-2507",
1798
- "qwen/qwen3-30b-a3b:free",
1799
- "qwen/qwen3-32b:free",
1800
- "qwen/qwen3-8b:free",
1801
- "qwen/qwen3-coder",
1802
- "qwen/qwen3-coder:free",
1803
- "qwen/qwen3-max",
1804
- "qwen/qwen3-next-80b-a3b-instruct",
1805
- "qwen/qwen3-next-80b-a3b-thinking",
1806
- "qwen/qwq-32b:free",
1807
- "rekaai/reka-flash-3",
1808
- "sarvamai/sarvam-m:free",
1809
- "thudm/glm-z1-32b:free",
1810
- "tngtech/deepseek-r1t2-chimera:free",
1811
- "x-ai/grok-3",
1812
- "x-ai/grok-3-beta",
1813
- "x-ai/grok-3-mini",
1814
- "x-ai/grok-3-mini-beta",
1815
- "x-ai/grok-4",
1816
- "x-ai/grok-4-fast",
1817
- "x-ai/grok-4-fast:free",
1818
- "x-ai/grok-code-fast-1",
1819
- "z-ai/glm-4.5",
1820
- "z-ai/glm-4.5-air",
1821
- "z-ai/glm-4.5-air:free",
1822
- "z-ai/glm-4.5v",
1823
- "z-ai/glm-4.6"
1778
+ anthropic: [
1779
+ "claude-3-5-haiku-20241022",
1780
+ "claude-3-5-haiku-latest",
1781
+ "claude-3-5-sonnet-20240620",
1782
+ "claude-3-5-sonnet-20241022",
1783
+ "claude-3-7-sonnet-20250219",
1784
+ "claude-3-7-sonnet-latest",
1785
+ "claude-3-haiku-20240307",
1786
+ "claude-3-opus-20240229",
1787
+ "claude-3-sonnet-20240229",
1788
+ "claude-haiku-4-5",
1789
+ "claude-haiku-4-5-20251001",
1790
+ "claude-opus-4-0",
1791
+ "claude-opus-4-1",
1792
+ "claude-opus-4-1-20250805",
1793
+ "claude-opus-4-20250514",
1794
+ "claude-sonnet-4-0",
1795
+ "claude-sonnet-4-20250514",
1796
+ "claude-sonnet-4-5",
1797
+ "claude-sonnet-4-5-20250929"
1824
1798
  ],
1825
- synthetic: [
1826
- "hf:Qwen/Qwen2.5-Coder-32B-Instruct",
1827
- "hf:Qwen/Qwen3-235B-A22B-Instruct-2507",
1828
- "hf:Qwen/Qwen3-235B-A22B-Thinking-2507",
1829
- "hf:Qwen/Qwen3-Coder-480B-A35B-Instruct",
1830
- "hf:deepseek-ai/DeepSeek-R1",
1831
- "hf:deepseek-ai/DeepSeek-R1-0528",
1832
- "hf:deepseek-ai/DeepSeek-V3",
1833
- "hf:deepseek-ai/DeepSeek-V3-0324",
1834
- "hf:deepseek-ai/DeepSeek-V3.1",
1835
- "hf:deepseek-ai/DeepSeek-V3.1-Terminus",
1836
- "hf:meta-llama/Llama-3.1-405B-Instruct",
1837
- "hf:meta-llama/Llama-3.1-70B-Instruct",
1838
- "hf:meta-llama/Llama-3.1-8B-Instruct",
1839
- "hf:meta-llama/Llama-3.3-70B-Instruct",
1840
- "hf:meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8",
1841
- "hf:meta-llama/Llama-4-Scout-17B-16E-Instruct",
1842
- "hf:moonshotai/Kimi-K2-Instruct",
1843
- "hf:moonshotai/Kimi-K2-Instruct-0905",
1844
- "hf:openai/gpt-oss-120b",
1845
- "hf:zai-org/GLM-4.5",
1846
- "hf:zai-org/GLM-4.6"
1799
+ moonshotai: ["kimi-k2-0711-preview", "kimi-k2-0905-preview", "kimi-k2-turbo-preview"],
1800
+ inference: [
1801
+ "google/gemma-3",
1802
+ "meta/llama-3.1-8b-instruct",
1803
+ "meta/llama-3.2-11b-vision-instruct",
1804
+ "meta/llama-3.2-1b-instruct",
1805
+ "meta/llama-3.2-3b-instruct",
1806
+ "mistral/mistral-nemo-12b-instruct",
1807
+ "osmosis/osmosis-structure-0.6b",
1808
+ "qwen/qwen-2.5-7b-vision-instruct",
1809
+ "qwen/qwen3-embedding-4b"
1810
+ ],
1811
+ baseten: ["Qwen3/Qwen3-Coder-480B-A35B-Instruct", "moonshotai/Kimi-K2-Instruct-0905", "zai-org/GLM-4.6"],
1812
+ cortecs: [
1813
+ "claude-4-5-sonnet",
1814
+ "claude-sonnet-4",
1815
+ "deepseek-v3-0324",
1816
+ "gemini-2.5-pro",
1817
+ "gpt-4.1",
1818
+ "gpt-oss-120b",
1819
+ "kimi-k2-instruct",
1820
+ "llama-3.1-405b-instruct",
1821
+ "nova-pro-v1",
1822
+ "qwen3-32b",
1823
+ "qwen3-coder-480b-a35b-instruct"
1824
+ ],
1825
+ mistral: [
1826
+ "codestral-latest",
1827
+ "devstral-medium-2507",
1828
+ "devstral-small-2505",
1829
+ "devstral-small-2507",
1830
+ "magistral-medium-latest",
1831
+ "magistral-small",
1832
+ "ministral-3b-latest",
1833
+ "ministral-8b-latest",
1834
+ "mistral-large-latest",
1835
+ "mistral-medium-2505",
1836
+ "mistral-medium-2508",
1837
+ "mistral-medium-latest",
1838
+ "mistral-nemo",
1839
+ "mistral-small-latest",
1840
+ "open-mistral-7b",
1841
+ "open-mixtral-8x22b",
1842
+ "open-mixtral-8x7b",
1843
+ "pixtral-12b",
1844
+ "pixtral-large-latest"
1845
+ ],
1846
+ venice: [
1847
+ "deepseek-coder-v2-lite",
1848
+ "deepseek-r1-671b",
1849
+ "dolphin-2.9.2-qwen2-72b",
1850
+ "llama-3.1-405b",
1851
+ "llama-3.2-3b",
1852
+ "llama-3.3-70b",
1853
+ "mistral-31-24b",
1854
+ "qwen-2.5-coder-32b",
1855
+ "qwen-2.5-qwq-32b",
1856
+ "qwen-2.5-vl",
1857
+ "qwen3-235b",
1858
+ "qwen3-4b",
1859
+ "venice-uncensored"
1860
+ ],
1861
+ alibaba: [
1862
+ "qvq-max",
1863
+ "qwen-flash",
1864
+ "qwen-max",
1865
+ "qwen-mt-plus",
1866
+ "qwen-mt-turbo",
1867
+ "qwen-omni-turbo",
1868
+ "qwen-omni-turbo-realtime",
1869
+ "qwen-plus",
1870
+ "qwen-plus-character-ja",
1871
+ "qwen-turbo",
1872
+ "qwen-vl-max",
1873
+ "qwen-vl-ocr",
1874
+ "qwen-vl-plus",
1875
+ "qwen2-5-14b-instruct",
1876
+ "qwen2-5-32b-instruct",
1877
+ "qwen2-5-72b-instruct",
1878
+ "qwen2-5-7b-instruct",
1879
+ "qwen2-5-omni-7b",
1880
+ "qwen2-5-vl-72b-instruct",
1881
+ "qwen2-5-vl-7b-instruct",
1882
+ "qwen3-14b",
1883
+ "qwen3-235b-a22b",
1884
+ "qwen3-32b",
1885
+ "qwen3-8b",
1886
+ "qwen3-asr-flash",
1887
+ "qwen3-coder-30b-a3b-instruct",
1888
+ "qwen3-coder-480b-a35b-instruct",
1889
+ "qwen3-coder-flash",
1890
+ "qwen3-coder-plus",
1891
+ "qwen3-livetranslate-flash-realtime",
1892
+ "qwen3-max",
1893
+ "qwen3-next-80b-a3b-instruct",
1894
+ "qwen3-next-80b-a3b-thinking",
1895
+ "qwen3-omni-flash",
1896
+ "qwen3-omni-flash-realtime",
1897
+ "qwen3-vl-235b-a22b",
1898
+ "qwen3-vl-30b-a3b",
1899
+ "qwen3-vl-plus",
1900
+ "qwq-plus"
1847
1901
  ],
1848
- deepinfra: [
1849
- "Qwen/Qwen3-Coder-480B-A35B-Instruct",
1850
- "Qwen/Qwen3-Coder-480B-A35B-Instruct-Turbo",
1851
- "moonshotai/Kimi-K2-Instruct",
1852
- "zai-org/GLM-4.5"
1902
+ "alibaba-cn": [
1903
+ "deepseek-r1",
1904
+ "deepseek-r1-0528",
1905
+ "deepseek-r1-distill-llama-70b",
1906
+ "deepseek-r1-distill-llama-8b",
1907
+ "deepseek-r1-distill-qwen-1-5b",
1908
+ "deepseek-r1-distill-qwen-14b",
1909
+ "deepseek-r1-distill-qwen-32b",
1910
+ "deepseek-r1-distill-qwen-7b",
1911
+ "deepseek-v3",
1912
+ "deepseek-v3-1",
1913
+ "deepseek-v3-2-exp",
1914
+ "moonshot-kimi-k2-instruct",
1915
+ "qvq-max",
1916
+ "qwen-deep-research",
1917
+ "qwen-doc-turbo",
1918
+ "qwen-flash",
1919
+ "qwen-long",
1920
+ "qwen-math-plus",
1921
+ "qwen-math-turbo",
1922
+ "qwen-max",
1923
+ "qwen-mt-plus",
1924
+ "qwen-mt-turbo",
1925
+ "qwen-omni-turbo",
1926
+ "qwen-omni-turbo-realtime",
1927
+ "qwen-plus",
1928
+ "qwen-plus-character",
1929
+ "qwen-turbo",
1930
+ "qwen-vl-max",
1931
+ "qwen-vl-ocr",
1932
+ "qwen-vl-plus",
1933
+ "qwen2-5-14b-instruct",
1934
+ "qwen2-5-32b-instruct",
1935
+ "qwen2-5-72b-instruct",
1936
+ "qwen2-5-7b-instruct",
1937
+ "qwen2-5-coder-32b-instruct",
1938
+ "qwen2-5-coder-7b-instruct",
1939
+ "qwen2-5-math-72b-instruct",
1940
+ "qwen2-5-math-7b-instruct",
1941
+ "qwen2-5-omni-7b",
1942
+ "qwen2-5-vl-72b-instruct",
1943
+ "qwen2-5-vl-7b-instruct",
1944
+ "qwen3-14b",
1945
+ "qwen3-235b-a22b",
1946
+ "qwen3-32b",
1947
+ "qwen3-8b",
1948
+ "qwen3-asr-flash",
1949
+ "qwen3-coder-30b-a3b-instruct",
1950
+ "qwen3-coder-480b-a35b-instruct",
1951
+ "qwen3-coder-flash",
1952
+ "qwen3-coder-plus",
1953
+ "qwen3-max",
1954
+ "qwen3-next-80b-a3b-instruct",
1955
+ "qwen3-next-80b-a3b-thinking",
1956
+ "qwen3-omni-flash",
1957
+ "qwen3-omni-flash-realtime",
1958
+ "qwen3-vl-235b-a22b",
1959
+ "qwen3-vl-30b-a3b",
1960
+ "qwen3-vl-plus",
1961
+ "qwq-32b",
1962
+ "qwq-plus",
1963
+ "tongyi-intent-detect-v3"
1853
1964
  ],
1854
- zhipuai: ["glm-4.5", "glm-4.5-air", "glm-4.5-flash", "glm-4.5v", "glm-4.6"],
1855
- submodel: [
1965
+ chutes: [
1856
1966
  "Qwen/Qwen3-235B-A22B-Instruct-2507",
1857
1967
  "Qwen/Qwen3-235B-A22B-Thinking-2507",
1968
+ "Qwen/Qwen3-30B-A3B",
1969
+ "Qwen/Qwen3-30B-A3B-Instruct-2507",
1970
+ "Qwen/Qwen3-30B-A3B-Thinking-2507",
1971
+ "Qwen/Qwen3-Coder-30B-A3B-Instruct",
1858
1972
  "Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8",
1973
+ "Qwen/Qwen3-Next-80B-A3B-Instruct",
1974
+ "Qwen/Qwen3-Next-80B-A3B-Thinking",
1975
+ "chutesai/Devstral-Small-2505",
1976
+ "chutesai/Mistral-Small-3.2-24B-Instruct-2506",
1859
1977
  "deepseek-ai/DeepSeek-R1-0528",
1978
+ "deepseek-ai/DeepSeek-R1-0528-Qwen3-8B",
1979
+ "deepseek-ai/DeepSeek-R1-Distill-Llama-70B",
1860
1980
  "deepseek-ai/DeepSeek-V3-0324",
1861
1981
  "deepseek-ai/DeepSeek-V3.1",
1982
+ "deepseek-ai/DeepSeek-V3.1-Terminus",
1983
+ "deepseek-ai/DeepSeek-V3.1-turbo",
1984
+ "deepseek-ai/DeepSeek-V3.1:THINKING",
1985
+ "deepseek-ai/DeepSeek-V3.2-Exp",
1986
+ "meituan-longcat/LongCat-Flash-Chat-FP8",
1987
+ "moonshotai/Kimi-Dev-72B",
1988
+ "moonshotai/Kimi-K2-Instruct-0905",
1989
+ "moonshotai/Kimi-K2-Instruct-75k",
1990
+ "moonshotai/Kimi-VL-A3B-Thinking",
1862
1991
  "openai/gpt-oss-120b",
1992
+ "tngtech/DeepSeek-R1T-Chimera",
1993
+ "tngtech/DeepSeek-TNG-R1T2-Chimera",
1863
1994
  "zai-org/GLM-4.5-Air",
1864
- "zai-org/GLM-4.5-FP8"
1865
- ],
1866
- zai: ["glm-4.5", "glm-4.5-air", "glm-4.5-flash", "glm-4.5v", "glm-4.6"],
1867
- inference: [
1868
- "google/gemma-3",
1869
- "meta/llama-3.1-8b-instruct",
1870
- "meta/llama-3.2-11b-vision-instruct",
1871
- "meta/llama-3.2-1b-instruct",
1872
- "meta/llama-3.2-3b-instruct",
1873
- "mistral/mistral-nemo-12b-instruct",
1874
- "osmosis/osmosis-structure-0.6b",
1875
- "qwen/qwen-2.5-7b-vision-instruct",
1876
- "qwen/qwen3-embedding-4b"
1877
- ],
1878
- requesty: [
1879
- "anthropic/claude-3-7-sonnet",
1880
- "anthropic/claude-4-sonnet-20250522",
1881
- "anthropic/claude-opus-4",
1882
- "anthropic/claude-opus-4-1-20250805",
1883
- "google/gemini-2.5-flash",
1884
- "google/gemini-2.5-pro",
1885
- "openai/gpt-4.1",
1886
- "openai/gpt-4.1-mini",
1887
- "openai/gpt-4o-mini",
1888
- "openai/gpt-5",
1889
- "openai/gpt-5-mini",
1890
- "openai/gpt-5-nano",
1891
- "openai/o4-mini"
1892
- ],
1893
- morph: ["auto", "morph-v3-fast", "morph-v3-large"],
1894
- lmstudio: ["openai/gpt-oss-20b", "qwen/qwen3-30b-a3b-2507", "qwen/qwen3-coder-30b"],
1895
- anthropic: [
1896
- "claude-3-5-haiku-20241022",
1897
- "claude-3-5-sonnet-20240620",
1898
- "claude-3-5-sonnet-20241022",
1899
- "claude-3-7-sonnet-20250219",
1900
- "claude-3-haiku-20240307",
1901
- "claude-3-opus-20240229",
1902
- "claude-3-sonnet-20240229",
1903
- "claude-haiku-4-5-20251001",
1904
- "claude-opus-4-1-20250805",
1905
- "claude-opus-4-20250514",
1906
- "claude-sonnet-4-20250514",
1907
- "claude-sonnet-4-5-20250929"
1908
- ],
1909
- "fireworks-ai": [
1910
- "accounts/fireworks/models/deepseek-r1-0528",
1911
- "accounts/fireworks/models/deepseek-v3-0324",
1912
- "accounts/fireworks/models/deepseek-v3p1",
1913
- "accounts/fireworks/models/glm-4p5",
1914
- "accounts/fireworks/models/glm-4p5-air",
1915
- "accounts/fireworks/models/gpt-oss-120b",
1916
- "accounts/fireworks/models/gpt-oss-20b",
1917
- "accounts/fireworks/models/kimi-k2-instruct",
1918
- "accounts/fireworks/models/qwen3-235b-a22b",
1919
- "accounts/fireworks/models/qwen3-coder-480b-a35b-instruct"
1920
- ],
1921
- modelscope: [
1922
- "Qwen/Qwen3-235B-A22B-Instruct-2507",
1923
- "Qwen/Qwen3-235B-A22B-Thinking-2507",
1924
- "Qwen/Qwen3-30B-A3B-Instruct-2507",
1925
- "Qwen/Qwen3-30B-A3B-Thinking-2507",
1926
- "Qwen/Qwen3-Coder-30B-A3B-Instruct",
1927
- "ZhipuAI/GLM-4.5"
1928
- ],
1929
- llama: [
1930
- "cerebras-llama-4-maverick-17b-128e-instruct",
1931
- "cerebras-llama-4-scout-17b-16e-instruct",
1932
- "groq-llama-4-maverick-17b-128e-instruct",
1933
- "llama-3.3-70b-instruct",
1934
- "llama-3.3-8b-instruct",
1935
- "llama-4-maverick-17b-128e-instruct-fp8",
1936
- "llama-4-scout-17b-16e-instruct-fp8"
1995
+ "zai-org/GLM-4.5-FP8",
1996
+ "zai-org/GLM-4.5-turbo",
1997
+ "zai-org/GLM-4.6-FP8",
1998
+ "zai-org/GLM-4.6-turbo"
1937
1999
  ],
1938
- cerebras: ["gpt-oss-120b", "qwen-3-235b-a22b-instruct-2507", "qwen-3-coder-480b"],
2000
+ upstage: ["solar-mini", "solar-pro2"],
1939
2001
  netlify: [
1940
2002
  "anthropic/claude-3-5-haiku-20241022",
1941
2003
  "anthropic/claude-3-5-haiku-latest",
@@ -2160,8 +2222,8 @@ var GatewayRegistry = class _GatewayRegistry {
2160
2222
  }
2161
2223
  this.isRefreshing = true;
2162
2224
  try {
2163
- const { ModelsDevGateway: ModelsDevGateway2 } = await import('./models-dev-3VEABUH4.js');
2164
- const { NetlifyGateway: NetlifyGateway2 } = await import('./netlify-NYFKA6UB.js');
2225
+ const { ModelsDevGateway: ModelsDevGateway2 } = await import('./models-dev-4VGIWYS3.js');
2226
+ const { NetlifyGateway: NetlifyGateway2 } = await import('./netlify-26MMY5X4.js');
2165
2227
  const { fetchProvidersFromGateways, writeRegistryFiles } = await import('./registry-generator-DXRSYYYT.js');
2166
2228
  const gateways2 = [new ModelsDevGateway2({}), new NetlifyGateway2()];
2167
2229
  const { providers, models } = await fetchProvidersFromGateways(gateways2);
@@ -2270,6 +2332,24 @@ if (autoRefreshEnabled) {
2270
2332
  GatewayRegistry.getInstance({ useDynamicLoading: isDev }).startAutoRefresh();
2271
2333
  }
2272
2334
 
2335
+ // src/llm/model/gateways/index.ts
2336
+ function findGatewayForModel(gatewayId, gateways2) {
2337
+ const prefixedGateway = gateways2.find((g) => g.prefix && gatewayId.startsWith(`${g.prefix}/`));
2338
+ if (prefixedGateway) {
2339
+ return prefixedGateway;
2340
+ }
2341
+ const unprefixedGateways = gateways2.filter((g) => !g.prefix);
2342
+ for (const gateway of unprefixedGateways) {
2343
+ return gateway;
2344
+ }
2345
+ throw new MastraError({
2346
+ id: "MODEL_ROUTER_NO_GATEWAY_FOUND",
2347
+ category: "USER",
2348
+ domain: "MODEL_ROUTER",
2349
+ text: `No Mastra model router gateway found for model id ${gatewayId}`
2350
+ });
2351
+ }
2352
+
2273
2353
  // src/llm/model/router.ts
2274
2354
  function getStaticProvidersByGateway(name) {
2275
2355
  return Object.fromEntries(Object.entries(PROVIDER_REGISTRY).filter(([_provider, config]) => config.gateway === name));
@@ -2325,7 +2405,11 @@ var ModelRouterLanguageModel = class _ModelRouterLanguageModel {
2325
2405
  async doStream(options) {
2326
2406
  let apiKey;
2327
2407
  try {
2328
- apiKey = this.config.apiKey || await this.gateway.getApiKey(this.config.routerId);
2408
+ if (this.config.url) {
2409
+ apiKey = this.config.apiKey || "";
2410
+ } else {
2411
+ apiKey = this.config.apiKey || await this.gateway.getApiKey(this.config.routerId);
2412
+ }
2329
2413
  } catch (error) {
2330
2414
  return {
2331
2415
  stream: new ReadableStream({
@@ -2356,7 +2440,8 @@ var ModelRouterLanguageModel = class _ModelRouterLanguageModel {
2356
2440
  name: providerId,
2357
2441
  apiKey,
2358
2442
  baseURL: this.config.url,
2359
- headers: this.config.headers
2443
+ headers: this.config.headers,
2444
+ supportsStructuredOutputs: true
2360
2445
  }).chatModel(modelId);
2361
2446
  _ModelRouterLanguageModel.modelInstances.set(key, modelInstance2);
2362
2447
  return modelInstance2;
@@ -2393,7 +2478,107 @@ async function resolveModelConfig(modelConfig, runtimeContext = new RuntimeConte
2393
2478
  }
2394
2479
  throw new Error("Invalid model configuration provided");
2395
2480
  }
2481
+ var ModelRouterEmbeddingModel = class {
2482
+ specificationVersion = "v2";
2483
+ modelId;
2484
+ provider;
2485
+ maxEmbeddingsPerCall = 2048;
2486
+ supportsParallelCalls = true;
2487
+ providerModel;
2488
+ constructor(config) {
2489
+ let normalizedConfig;
2490
+ if (typeof config === "string") {
2491
+ const parts = config.split("/");
2492
+ if (parts.length !== 2) {
2493
+ throw new Error(`Invalid model string format: "${config}". Expected format: "provider/model"`);
2494
+ }
2495
+ const [providerId, modelId] = parts;
2496
+ normalizedConfig = { providerId, modelId };
2497
+ } else if ("providerId" in config && "modelId" in config) {
2498
+ normalizedConfig = {
2499
+ providerId: config.providerId,
2500
+ modelId: config.modelId,
2501
+ url: config.url,
2502
+ apiKey: config.apiKey,
2503
+ headers: config.headers
2504
+ };
2505
+ } else {
2506
+ const parts = config.id.split("/");
2507
+ if (parts.length !== 2) {
2508
+ throw new Error(`Invalid model string format: "${config.id}". Expected format: "provider/model"`);
2509
+ }
2510
+ const [providerId, modelId] = parts;
2511
+ normalizedConfig = {
2512
+ providerId,
2513
+ modelId,
2514
+ url: config.url,
2515
+ apiKey: config.apiKey,
2516
+ headers: config.headers
2517
+ };
2518
+ }
2519
+ this.provider = normalizedConfig.providerId;
2520
+ this.modelId = normalizedConfig.modelId;
2521
+ if (normalizedConfig.url) {
2522
+ const apiKey = normalizedConfig.apiKey || "";
2523
+ this.providerModel = createOpenAICompatible({
2524
+ name: normalizedConfig.providerId,
2525
+ apiKey,
2526
+ baseURL: normalizedConfig.url,
2527
+ headers: normalizedConfig.headers
2528
+ }).textEmbeddingModel(normalizedConfig.modelId);
2529
+ } else {
2530
+ const registry = GatewayRegistry.getInstance();
2531
+ const providerConfig = registry.getProviderConfig(normalizedConfig.providerId);
2532
+ if (!providerConfig) {
2533
+ throw new Error(`Unknown provider: ${normalizedConfig.providerId}`);
2534
+ }
2535
+ let apiKey = normalizedConfig.apiKey;
2536
+ if (!apiKey) {
2537
+ const apiKeyEnvVar = providerConfig.apiKeyEnvVar;
2538
+ if (Array.isArray(apiKeyEnvVar)) {
2539
+ for (const envVar of apiKeyEnvVar) {
2540
+ apiKey = process.env[envVar];
2541
+ if (apiKey) break;
2542
+ }
2543
+ } else {
2544
+ apiKey = process.env[apiKeyEnvVar];
2545
+ }
2546
+ }
2547
+ if (!apiKey) {
2548
+ const envVarDisplay = Array.isArray(providerConfig.apiKeyEnvVar) ? providerConfig.apiKeyEnvVar.join(" or ") : providerConfig.apiKeyEnvVar;
2549
+ throw new Error(`API key not found for provider ${normalizedConfig.providerId}. Set ${envVarDisplay}`);
2550
+ }
2551
+ if (normalizedConfig.providerId === "openai") {
2552
+ this.providerModel = createOpenAI({ apiKey }).textEmbeddingModel(
2553
+ normalizedConfig.modelId
2554
+ );
2555
+ } else if (normalizedConfig.providerId === "google") {
2556
+ this.providerModel = createGoogleGenerativeAI({ apiKey }).textEmbedding(
2557
+ normalizedConfig.modelId
2558
+ );
2559
+ } else {
2560
+ if (!providerConfig.url) {
2561
+ throw new Error(`Provider ${normalizedConfig.providerId} does not have a URL configured`);
2562
+ }
2563
+ this.providerModel = createOpenAICompatible({
2564
+ name: normalizedConfig.providerId,
2565
+ apiKey,
2566
+ baseURL: providerConfig.url
2567
+ }).textEmbeddingModel(normalizedConfig.modelId);
2568
+ }
2569
+ }
2570
+ if (this.providerModel.maxEmbeddingsPerCall !== void 0) {
2571
+ this.maxEmbeddingsPerCall = this.providerModel.maxEmbeddingsPerCall;
2572
+ }
2573
+ if (this.providerModel.supportsParallelCalls !== void 0) {
2574
+ this.supportsParallelCalls = this.providerModel.supportsParallelCalls;
2575
+ }
2576
+ }
2577
+ async doEmbed(args) {
2578
+ return this.providerModel.doEmbed(args);
2579
+ }
2580
+ };
2396
2581
 
2397
- export { ModelRouterLanguageModel, PROVIDER_REGISTRY, getProviderConfig, parseModelString, resolveModelConfig };
2398
- //# sourceMappingURL=chunk-HP7B2TQQ.js.map
2399
- //# sourceMappingURL=chunk-HP7B2TQQ.js.map
2582
+ export { ModelRouterEmbeddingModel, ModelRouterLanguageModel, PROVIDER_REGISTRY, getProviderConfig, parseModelString, resolveModelConfig };
2583
+ //# sourceMappingURL=chunk-NIESGYOH.js.map
2584
+ //# sourceMappingURL=chunk-NIESGYOH.js.map