chore: update model catalog from bot issues#1000
Conversation
|
Codex (@codex) review |
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 30b332db65
ℹ️ About Codex in GitHub
Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "Codex (@codex) review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "Codex (@codex) address that feedback".
| "publishers/openai/models/gpt-oss-120b-maas": ["vertex"], | ||
| "publishers/openai/models/gpt-oss-20b-maas": ["vertex"], | ||
| "publishers/xai/models/grok-4.3": ["vertex"], | ||
| "publishers/xai/models/grok-4.20-non-reasoning": ["vertex"], |
There was a problem hiding this comment.
When these new OpenAI/xAI Vertex models are selected, packages/proxy/src/proxy.ts:2202-2225 only sends publishers/meta and publishers/qwen through the OpenAI-compatible /endpoints/openapi/chat/completions path; these entries therefore fall into the rawPredict branch. Google/xAI docs for Grok use the OpenAI-compatible endpoint and xai/... model IDs, and Google's gpt-oss MaaS examples likewise use the OpenAI-compatible API with openai/... IDs, so these newly advertised Vertex options will return not-found/unsupported until the routing and model-id rewrite are extended. Sources: https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/partner-models/grok/capabilities/function-calling, https://discuss.google.dev/t/now-ga-openais-gpt-oss-qwen3-models-on-vertex-ai-as-open-model-apis/253945
Useful? React with 👍 / 👎.
| "available_providers": [ | ||
| "baseten" | ||
| "baseten", | ||
| "together" |
There was a problem hiding this comment.
For Together requests the proxy forwards bodyData.model unchanged to the OpenAI-compatible endpoint, and a repo-wide search only finds this lowercase thinkingmachines/inkling catalog key. Together's serverless catalog lists the API model string as thinkingmachines/Inkling, so enabling Together on the lowercase Baseten id advertises a model name that Together users are likely to have rejected; add a correct-case Together entry or a provider-specific translation instead. Source: https://docs.together.ai/docs/serverless/models
Useful? React with 👍 / 👎.
1 participant
Automated daily batch of model catalog updates from bot issues.
Included issues
Summary
grok-4.5grok-4.5-latestgrok-4.5grok-4.5-latest2
gpt-realtime-2.1gpt-realtime-2.1-minigpt-realtime-2.1gpt-realtime-2.1-mini2
3
4
5
gpt-4oo1o3-minigpt-4oo1o3-mini2
o4-minigpt-4-turbogpt-4.1-nanoo4-minigpt-4-turbogpt-4.1-nanopublishers/mistralai/models/mistral-medium-3publishers/mistralai/models/mistral-small-2503publishers/mistralai/models/codestral-2publishers/mistralai/models/mistral-medium-3publishers/mistralai/models/mistral-small-2503publishers/mistralai/models/codestral-22
3
4
5
6
publishers/openai/models/gpt-oss-120b-maaspublishers/openai/models/gpt-oss-20b-maaspublishers/openai/models/gpt-oss-120b-maaspublishers/openai/models/gpt-oss-20b-maas2
3
4
5
publishers/xai/models/grok-4.3publishers/xai/models/grok-4.20-non-reasoningpublishers/xai/models/grok-4.3publishers/xai/models/grok-4.20-non-reasoning2
3
thinkingmachines/inklingthinkingmachines/inkling2
Verified metadata
#990: [BOT ISSUE] xAI: fix stale cached input pricing for grok-4.5 and grok-4.5-latest ($0.50 → $0.30)
Verification notes
No LLM verification step ran; model metadata was already complete in the issue.
sync_models vs proposed update
sync_models cross-check found differences. Official provider verification was used for the applied values, and sync_models discrepancies are listed below for review.
#991: [BOT ISSUE] OpenAI: add missing gpt-realtime-2.1 and gpt-realtime-2.1-mini
Verification notes
Verification
sync_models vs proposed update
sync_models cross-check found differences. Official provider verification was used for the applied values, and sync_models discrepancies are listed below for review.
#992: [BOT ISSUE] OpenAI: add deprecation_date for gpt-4o, o1, o3-mini (shutdown 2026-10-23)
Verification notes
Verification
sync_models vs proposed update
sync_models cross-check found differences. Official provider verification was used for the applied values, and sync_models discrepancies are listed below for review.
#993: [BOT ISSUE] OpenAI: add deprecation_date for o4-mini, gpt-4-turbo, gpt-4.1-nano (shutdown 2026-10-23)
Verification notes
Verification
sync_models vs proposed update
sync_models cross-check found differences. Official provider verification was used for the applied values, and sync_models discrepancies are listed below for review.
#994: [BOT ISSUE] Vertex: add missing Mistral MaaS entries (mistral-medium-3, mistral-small-2503, codestral-2)
Verification notes
Verification
sync_models vs proposed update
sync_models cross-check found differences. Official provider verification was used for the applied values, and sync_models discrepancies are listed below for review.
#996: [BOT ISSUE] Vertex: add missing OpenAI GPT-OSS MaaS entries (gpt-oss-120b-maas, gpt-oss-20b-maas)
Verification notes
Verification
sync_models vs proposed update
sync_models cross-check found differences. Official provider verification was used for the applied values, and sync_models discrepancies are listed below for review.
#997: [BOT ISSUE] Vertex: add missing xAI Grok MaaS entries (grok-4.3, grok-4.20-non-reasoning)
Verification notes
Verification
sync_models vs proposed update
sync_models cross-check found differences. Official provider verification was used for the applied values, and sync_models discrepancies are listed below for review.
#999: [BOT ISSUE] Together: add together to available_providers for thinkingmachines/inkling
Verification notes
Verification
sync_models vs proposed update
sync_models cross-check found differences. Official provider verification was used for the applied values, and sync_models discrepancies are listed below for review.