@magpie-community/opencode-factory-auth 0.1.38 → 0.1.39
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.
- package/README.md +11 -2
- package/index.mjs +1 -0
- package/package.json +1 -1
package/README.md
CHANGED
|
@@ -65,14 +65,23 @@ Each model is served on the one API droid uses for it:
|
|
|
65
65
|
|---|---|
|
|
66
66
|
| Anthropic Messages (`/api/llm/a`) | Fable 5.1, Fable 5, Opus 5.5, Opus 5, Opus 4.8 (each also in Fast Mode), Sonnet 5.5, Sonnet 5, Sonnet 4.6, Haiku 5.5, Haiku 4.5, MiniMax M2.7 |
|
|
67
67
|
| OpenAI Responses (`/api/llm/o/v1`) | GPT-6.1 Sol, GPT-6 Sol/Astra/Luna, GPT-5.6 Sol/Terra/Luna, GPT-5.5, GPT-5.4, GPT-5.3-Codex (each also in Fast Mode), Grok 4.7, Grok 4.6 |
|
|
68
|
-
| Chat completions (`/api/llm/o/v1`) | GLM-5.3, GLM-5.3-Flash, GLM-5.2, Kimi K3, DeepSeek V4.1 Flash, Qwen3.8 Max, MiniMax M3, Mistral Large 4, Mistral Medium 3.5, Nemotron 3 Ultra |
|
|
68
|
+
| Chat completions (`/api/llm/o/v1`) | GLM-5.3, GLM-5.3-Flash, GLM-5.2, Kimi K3, DeepSeek V4.1 Flash, DeepSeek V4.1 Flash 1017, Qwen3.8 Max, MiniMax M3, Mistral Large 4, Mistral Medium 3.5, Nemotron 3 Ultra |
|
|
69
69
|
| Gemini generate (`/api/llm/g/v1/generate`) | Gemini 4 Argon, Gemini 3.1 Pro, Gemini 3.8/3.7/3.6/3.5 Flash, Gemini 3 Flash |
|
|
70
70
|
|
|
71
71
|
Reasoning efforts are the variants droid offers for each model. DeepSeek V4.1
|
|
72
|
-
Flash, Kimi K3 and GLM-5.2 have `none`, droid's "off": they stop
|
|
72
|
+
Flash (1017 too), Kimi K3 and GLM-5.2 have `none`, droid's "off": they stop
|
|
73
|
+
thinking.
|
|
73
74
|
Reasoning is mandatory on GLM-5.3 and GLM-5.3-Flash (Fireworks turns `none`
|
|
74
75
|
away), so `low` is their least.
|
|
75
76
|
|
|
77
|
+
DeepSeek V4.1 Flash 1017 (`deepseek-v4.1-flash-1017`) is a separate model
|
|
78
|
+
entry, with droid 0.251.0's 1,040,000-token context, 131,072-token output,
|
|
79
|
+
image support, `none`/`low`/`high`/`max` efforts and 0.12x token multiplier.
|
|
80
|
+
Like every other model here it goes to its registry's default upstream,
|
|
81
|
+
`apiProviders[0]`, which is Fireworks, as droid sends it with
|
|
82
|
+
`x-provider-routing-source: registry_default`. Factory gates 1017 with a
|
|
83
|
+
feature flag, so availability still depends on the account.
|
|
84
|
+
|
|
76
85
|
A model's price is its token multiplier in droid's registry
|
|
77
86
|
(`cost.tokenMultiplier`), which magpie shows beside the model as its rate
|
|
78
87
|
(`0.8×`), so a cheap one is picked without opening droid. A promotion running
|
package/index.mjs
CHANGED
|
@@ -87,6 +87,7 @@ const MODELS = [
|
|
|
87
87
|
["glm-5.2", "GLM-5.2", CHAT, "baseten", 1040000, 131072, ["none", "high", "max"], false, 0.56],
|
|
88
88
|
["kimi-k3", "Kimi K3", CHAT, "fireworks", 262144, 65536, ["none", "low", "high", "max"], true, 1.2],
|
|
89
89
|
["deepseek-v4.1-flash", "DeepSeek V4.1 Flash", CHAT, "fireworks", 1040000, 131072, ["none", "low", "high", "max"], true, 0.12],
|
|
90
|
+
["deepseek-v4.1-flash-1017", "DeepSeek V4.1 Flash 1017", CHAT, "fireworks", 1040000, 131072, ["none", "low", "high", "max"], true, 0.12],
|
|
90
91
|
["qwen3.8-max", "Qwen3.8 Max", CHAT, "fireworks", 262144, 131072, ["low", "medium", "xhigh"], false, 0.8],
|
|
91
92
|
["minimax-m3", "MiniMax M3", CHAT, "fireworks", 512000, 64000, ["high"], true, 0.12],
|
|
92
93
|
["minimax-m2.7", "MiniMax M2.7", ANTHROPIC, "fireworks", 196600, 64000, ["high"], false, 0.12],
|