Default Kimi K3 to max reasoning effort

[chrarnoldus]

Summary

• add max, high, and low reasoning variants in that order
• map kimi-k3 to the new variant set so max is the first/default option
• keep kimi-k2.7-code on the existing thinking-only setting

Verification

• pnpm --filter web test -- --runInBand apps/web/src/lib/ai-gateway/custom-llm/order-opencode-variants.test.ts apps/web/src/lib/ai-gateway/providers/direct-byok/opencode-go.test.ts
• pnpm --filter web typecheck
• pnpm --filter web lint
• git diff --check

[kilo-code-bot]

Code Review Summary

Status: No Issues Found | Recommendation: Merge

Executive Summary

The new REASONING_VARIANTS_MAX_HIGH_LOW variant set and its kimi-k3 mapping are self-contained and correctly ordered (max, high, low), 'max' is a valid ReasoningEffortSchema value, and splitting kimi-k3 from kimi-k2.7-code does not affect any other model-matching branch or downstream variant consumer.

Files Reviewed (1 files)

• apps/web/src/lib/ai-gateway/providers/model-settings.ts

---

_{Reviewed by claude-sonnet-5 · Input: 24 · Output: 4.9K · Cached: 474.6K}

_{Review guidance: REVIEW.md from base branch main}