Repository navigation
chore: add model to catalog - #1719
github-actions[bot] wants to merge 1 commit into
Conversation
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
![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: 951671858f
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".
| "format": "openai", | ||
| "flavor": "chat", | ||
| "displayName": "moonshotai-kimi-k3", | ||
| "max_input_tokens": 8192, |
There was a problem hiding this comment.
For Cerebras users sending prompts above 8,192 tokens, this catalog entry incorrectly reports that the request exceeds the model's input capacity. The other Kimi K3 entries in this catalog consistently declare a 1,000,000 or 1,048,576-token input limit, including the Fireworks entry at lines 7751-7761 and the OpenRouter entry at lines 27208-27218. Replace this apparent placeholder limit with the documented limit for the Cerebras deployment so catalog consumers do not reject or truncate valid long-context requests.
Useful? React with 👍 / 👎.
| "format": "openai", | ||
| "flavor": "chat", | ||
| "displayName": "moonshotai-kimi-k3", |
There was a problem hiding this comment.
When clients use the catalog to decide whether image inputs and reasoning controls are supported, this new alias exposes both capabilities as falsey because it omits multimodal and reasoning. Kimi K3 is a multimodal reasoning model, and the existing Fireworks, Bedrock, Databricks, and OpenRouter K3 entries all set both fields to true; add the same capability metadata here so choosing the Cerebras alias does not hide supported inputs and controls.
Useful? React with 👍 / 👎.
Automated targeted model-catalog addition from the model-drop workflow.