Skip to content

[Docs] Add Kimi Code provider page with reasoning effort setting#17

Draft
zoomote[bot] wants to merge 1 commit into
mainfrom
docs/kimi-code-provider-2aq5n5g050jl4
Draft

[Docs] Add Kimi Code provider page with reasoning effort setting#17
zoomote[bot] wants to merge 1 commit into
mainfrom
docs/kimi-code-provider-2aq5n5g050jl4

Conversation

@zoomote

@zoomote zoomote Bot commented Jul 23, 2026

Copy link
Copy Markdown

Opened on behalf of Toray Altas. Follow up by mentioning @zoomote, in the web UI, or in Discord.

What changed

Adds a new Kimi Code provider documentation page (docs/providers/kimi-code.md) and registers it in the Model Providers navigation (docs/providers/index.json). The page covers OAuth device-flow and API-key authentication, automatic model discovery, and the configurable Model Reasoning Effort setting (Low / High / Max, default Max) that was added to the Kimi Code provider in Zoo-Code-Org/Zoo-Code#945.

Why this change was made

The Kimi Code provider is new and had no documentation. Zoo Code's convention is that functionality changes ship with a paired docs update; the reasoning effort dropdown is a user-visible setting that needed an explanation, including the fact that Kimi K3 always thinks (no "None" option).

Impact

Users configuring the Kimi Code provider can now find setup steps, the reasoning effort options, and common issues in the docs. No product behavior changes.

Screenshots

New Kimi Code provider docs page rendering the Configuration > Reasoning Effort section with Low / High / Max (default Max)

Model Providers sidebar showing the new Kimi Code entry between Mistral AI and Moonshot

Related PRs

@zoomote zoomote Bot assigned taltas Jul 23, 2026
@zoomote

zoomote Bot commented Jul 23, 2026

Copy link
Copy Markdown
Author

No code issues found. See task

Reviewed the new Kimi Code provider page and navigation entry against the paired implementation in Zoo-Code-Org/Zoo-Code#945 (feat/kimi-code branch). Verified claims: reasoning effort Low/High/Max with Max default and no "None" option (requiredReasoningEffort), 32,768 max output tokens, OAuth device flow with single automatic 401 retry, the "Not authenticated with Kimi Code" error, "Model Reasoning Effort" and "Refresh Models" settings UI labels, and the https://www.kimi.com/code link. index.json entry wires correctly into the sidebar via sidebars.ts, the page follows existing provider-page conventions (e.g. qwen-code.md), and pnpm run check-types passes.

Reviewed 92a2df0

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant