@earendil-works/pi-ai 0.79.3 → 0.79.4

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.
@@ -4191,7 +4191,7 @@ export const MODELS = {
4191
4191
  headers: { "User-Agent": "GitHubCopilotChat/0.35.0", "Editor-Version": "vscode/1.107.0", "Editor-Plugin-Version": "copilot-chat/0.35.0", "Copilot-Integration-Id": "vscode-chat" },
4192
4192
  compat: { "forceAdaptiveThinking": true, "supportsTemperature": false },
4193
4193
  reasoning: true,
4194
- thinkingLevelMap: { "xhigh": "xhigh" },
4194
+ thinkingLevelMap: { "xhigh": "xhigh", "minimal": "low" },
4195
4195
  input: ["text", "image"],
4196
4196
  cost: {
4197
4197
  input: 5,
@@ -4211,7 +4211,7 @@ export const MODELS = {
4211
4211
  headers: { "User-Agent": "GitHubCopilotChat/0.35.0", "Editor-Version": "vscode/1.107.0", "Editor-Plugin-Version": "copilot-chat/0.35.0", "Copilot-Integration-Id": "vscode-chat" },
4212
4212
  compat: { "forceAdaptiveThinking": true, "supportsTemperature": false },
4213
4213
  reasoning: true,
4214
- thinkingLevelMap: { "xhigh": "xhigh" },
4214
+ thinkingLevelMap: { "xhigh": "xhigh", "minimal": "low" },
4215
4215
  input: ["text", "image"],
4216
4216
  cost: {
4217
4217
  input: 5,
@@ -4269,6 +4269,7 @@ export const MODELS = {
4269
4269
  headers: { "User-Agent": "GitHubCopilotChat/0.35.0", "Editor-Version": "vscode/1.107.0", "Editor-Plugin-Version": "copilot-chat/0.35.0", "Copilot-Integration-Id": "vscode-chat" },
4270
4270
  compat: { "forceAdaptiveThinking": true },
4271
4271
  reasoning: true,
4272
+ thinkingLevelMap: { "minimal": "low", "xhigh": "max" },
4272
4273
  input: ["text", "image"],
4273
4274
  cost: {
4274
4275
  input: 3,
@@ -4828,6 +4829,42 @@ export const MODELS = {
4828
4829
  contextWindow: 262144,
4829
4830
  maxTokens: 32768,
4830
4831
  },
4832
+ "gemma-4-E2B-it": {
4833
+ id: "gemma-4-E2B-it",
4834
+ name: "Gemma 4 E2B IT",
4835
+ api: "google-generative-ai",
4836
+ provider: "google",
4837
+ baseUrl: "https://generativelanguage.googleapis.com/v1beta",
4838
+ reasoning: true,
4839
+ thinkingLevelMap: { "off": null, "minimal": "MINIMAL", "low": null, "medium": null, "high": "HIGH" },
4840
+ input: ["text", "image"],
4841
+ cost: {
4842
+ input: 0,
4843
+ output: 0,
4844
+ cacheRead: 0,
4845
+ cacheWrite: 0,
4846
+ },
4847
+ contextWindow: 131072,
4848
+ maxTokens: 8192,
4849
+ },
4850
+ "gemma-4-E4B-it": {
4851
+ id: "gemma-4-E4B-it",
4852
+ name: "Gemma 4 E4B IT",
4853
+ api: "google-generative-ai",
4854
+ provider: "google",
4855
+ baseUrl: "https://generativelanguage.googleapis.com/v1beta",
4856
+ reasoning: true,
4857
+ thinkingLevelMap: { "off": null, "minimal": "MINIMAL", "low": null, "medium": null, "high": "HIGH" },
4858
+ input: ["text", "image"],
4859
+ cost: {
4860
+ input: 0,
4861
+ output: 0,
4862
+ cacheRead: 0,
4863
+ cacheWrite: 0,
4864
+ },
4865
+ contextWindow: 131072,
4866
+ maxTokens: 8192,
4867
+ },
4831
4868
  },
4832
4869
  "google-vertex": {
4833
4870
  "gemini-1.5-flash": {
@@ -7886,7 +7923,7 @@ export const MODELS = {
7886
7923
  api: "openai-completions",
7887
7924
  provider: "opencode",
7888
7925
  baseUrl: "https://opencode.ai/zen/v1",
7889
- compat: { "maxTokensField": "max_tokens", "requiresReasoningContentOnAssistantMessages": true, "thinkingFormat": "deepseek" },
7926
+ compat: { "maxTokensField": "max_tokens", "supportsLongCacheRetention": false, "requiresReasoningContentOnAssistantMessages": true, "thinkingFormat": "deepseek" },
7890
7927
  reasoning: true,
7891
7928
  thinkingLevelMap: { "minimal": null, "low": null, "medium": null, "high": "high", "xhigh": "max" },
7892
7929
  input: ["text"],
@@ -7924,7 +7961,7 @@ export const MODELS = {
7924
7961
  api: "openai-completions",
7925
7962
  provider: "opencode",
7926
7963
  baseUrl: "https://opencode.ai/zen/v1",
7927
- compat: { "maxTokensField": "max_tokens", "requiresReasoningContentOnAssistantMessages": true, "thinkingFormat": "deepseek" },
7964
+ compat: { "maxTokensField": "max_tokens", "supportsLongCacheRetention": false, "requiresReasoningContentOnAssistantMessages": true, "thinkingFormat": "deepseek" },
7928
7965
  reasoning: true,
7929
7966
  thinkingLevelMap: { "minimal": null, "low": null, "medium": null, "high": "high", "xhigh": "max" },
7930
7967
  input: ["text"],
@@ -8340,7 +8377,7 @@ export const MODELS = {
8340
8377
  api: "openai-completions",
8341
8378
  provider: "opencode",
8342
8379
  baseUrl: "https://opencode.ai/zen/v1",
8343
- compat: { "maxTokensField": "max_tokens" },
8380
+ compat: { "maxTokensField": "max_tokens", "supportsLongCacheRetention": false },
8344
8381
  reasoning: true,
8345
8382
  input: ["text", "image"],
8346
8383
  cost: {
@@ -8358,7 +8395,7 @@ export const MODELS = {
8358
8395
  api: "openai-completions",
8359
8396
  provider: "opencode",
8360
8397
  baseUrl: "https://opencode.ai/zen/v1",
8361
- compat: { "thinkingFormat": "deepseek", "supportsReasoningEffort": false, "maxTokensField": "max_tokens" },
8398
+ compat: { "thinkingFormat": "deepseek", "supportsReasoningEffort": false, "maxTokensField": "max_tokens", "supportsLongCacheRetention": false },
8362
8399
  reasoning: true,
8363
8400
  input: ["text", "image"],
8364
8401
  cost: {
@@ -8412,7 +8449,7 @@ export const MODELS = {
8412
8449
  api: "openai-completions",
8413
8450
  provider: "opencode",
8414
8451
  baseUrl: "https://opencode.ai/zen/v1",
8415
- compat: { "maxTokensField": "max_tokens" },
8452
+ compat: { "maxTokensField": "max_tokens", "supportsLongCacheRetention": false },
8416
8453
  reasoning: true,
8417
8454
  input: ["text"],
8418
8455
  cost: {
@@ -8576,7 +8613,7 @@ export const MODELS = {
8576
8613
  api: "openai-completions",
8577
8614
  provider: "opencode-go",
8578
8615
  baseUrl: "https://opencode.ai/zen/go/v1",
8579
- compat: { "thinkingFormat": "deepseek", "supportsReasoningEffort": false, "maxTokensField": "max_tokens" },
8616
+ compat: { "thinkingFormat": "deepseek", "supportsReasoningEffort": false, "maxTokensField": "max_tokens", "supportsLongCacheRetention": false },
8580
8617
  reasoning: true,
8581
8618
  thinkingLevelMap: { "minimal": null, "low": null, "medium": null },
8582
8619
  input: ["text", "image"],
@@ -9429,13 +9466,13 @@ export const MODELS = {
9429
9466
  thinkingLevelMap: { "minimal": null, "low": null, "medium": null, "high": "high", "xhigh": "xhigh" },
9430
9467
  input: ["text"],
9431
9468
  cost: {
9432
- input: 0.098,
9433
- output: 0.196,
9469
+ input: 0.09,
9470
+ output: 0.18,
9434
9471
  cacheRead: 0.02,
9435
9472
  cacheWrite: 0,
9436
9473
  },
9437
9474
  contextWindow: 1048576,
9438
- maxTokens: 4096,
9475
+ maxTokens: 65536,
9439
9476
  },
9440
9477
  "deepseek/deepseek-v4-pro": {
9441
9478
  id: "deepseek/deepseek-v4-pro",
@@ -10470,13 +10507,13 @@ export const MODELS = {
10470
10507
  reasoning: true,
10471
10508
  input: ["text", "image"],
10472
10509
  cost: {
10473
- input: 0.95,
10474
- output: 4,
10475
- cacheRead: 0.19,
10510
+ input: 0.75,
10511
+ output: 3.5,
10512
+ cacheRead: 0.16,
10476
10513
  cacheWrite: 0,
10477
10514
  },
10478
10515
  contextWindow: 262144,
10479
- maxTokens: 4096,
10516
+ maxTokens: 262144,
10480
10517
  },
10481
10518
  "nex-agi/nex-n2-pro:free": {
10482
10519
  id: "nex-agi/nex-n2-pro:free",
@@ -13301,6 +13338,25 @@ export const MODELS = {
13301
13338
  contextWindow: 262144,
13302
13339
  maxTokens: 131000,
13303
13340
  },
13341
+ "moonshotai/Kimi-K2.7-Code": {
13342
+ id: "moonshotai/Kimi-K2.7-Code",
13343
+ name: "Kimi K2.7 Code",
13344
+ api: "openai-completions",
13345
+ provider: "together",
13346
+ baseUrl: "https://api.together.ai/v1",
13347
+ compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "maxTokensField": "max_tokens", "supportsStrictMode": false, "supportsLongCacheRetention": false, "thinkingFormat": "together" },
13348
+ reasoning: true,
13349
+ thinkingLevelMap: { "minimal": null, "low": null, "medium": null },
13350
+ input: ["text"],
13351
+ cost: {
13352
+ input: 0.95,
13353
+ output: 4,
13354
+ cacheRead: 0.19,
13355
+ cacheWrite: 0,
13356
+ },
13357
+ contextWindow: 262144,
13358
+ maxTokens: 131072,
13359
+ },
13304
13360
  "nvidia/nemotron-3-ultra-550b-a55b": {
13305
13361
  id: "nvidia/nemotron-3-ultra-550b-a55b",
13306
13362
  name: "Nemotron 3 Ultra 550B A55B",
@@ -16820,6 +16876,24 @@ export const MODELS = {
16820
16876
  contextWindow: 200000,
16821
16877
  maxTokens: 131072,
16822
16878
  },
16879
+ "glm-5.2": {
16880
+ id: "glm-5.2",
16881
+ name: "GLM-5.2",
16882
+ api: "openai-completions",
16883
+ provider: "zai",
16884
+ baseUrl: "https://api.z.ai/api/coding/paas/v4",
16885
+ compat: { "supportsDeveloperRole": false, "thinkingFormat": "zai", "zaiToolStream": true },
16886
+ reasoning: true,
16887
+ input: ["text"],
16888
+ cost: {
16889
+ input: 0,
16890
+ output: 0,
16891
+ cacheRead: 0,
16892
+ cacheWrite: 0,
16893
+ },
16894
+ contextWindow: 1000000,
16895
+ maxTokens: 131072,
16896
+ },
16823
16897
  "glm-5v-turbo": {
16824
16898
  id: "glm-5v-turbo",
16825
16899
  name: "GLM-5V-Turbo",
@@ -16912,6 +16986,24 @@ export const MODELS = {
16912
16986
  contextWindow: 200000,
16913
16987
  maxTokens: 131072,
16914
16988
  },
16989
+ "glm-5.2": {
16990
+ id: "glm-5.2",
16991
+ name: "GLM-5.2",
16992
+ api: "openai-completions",
16993
+ provider: "zai-coding-cn",
16994
+ baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4",
16995
+ compat: { "supportsDeveloperRole": false, "thinkingFormat": "zai", "zaiToolStream": true },
16996
+ reasoning: true,
16997
+ input: ["text"],
16998
+ cost: {
16999
+ input: 0,
17000
+ output: 0,
17001
+ cacheRead: 0,
17002
+ cacheWrite: 0,
17003
+ },
17004
+ contextWindow: 1000000,
17005
+ maxTokens: 131072,
17006
+ },
16915
17007
  "glm-5v-turbo": {
16916
17008
  id: "glm-5v-turbo",
16917
17009
  name: "GLM-5V-Turbo",