Skip to content

Add Space Bunny Alpha to Auto Free and remove delisted Ling 3.0 Flash VL - #6716

Merged
chrarnoldus merged 3 commits into
mainfrom
chore/auto-free-space-bunny
Sep 25, 2026
Merged

chrarnoldus merged 3 commits into
mainfrom
chore/auto-free-space-bunny

Conversation

@chrarnoldus

@chrarnoldus chrarnoldus commented Sep 25, 2026 •

Copy link
Copy Markdown
Contributor

Summary

Updates the kilo-auto/free candidate pool in apps/web/src/lib/ai-gateway/models.ts:

  • Removed inclusionai/ling-3.0-flash-vl:free: OpenRouter's /api/v1/models no longer lists it. The other existing models (poolside/laguna-s-2.1:free, nvidia/nemotron-3-ultra-550b-a55b:free, dots-studio/dots-3-note-preview:free, nex-agi/nex-n2.5-pro:free) are still listed and stay in the pool.
  • Added stealth/space-bunny-alpha (Space Bunny Alpha). It is listed on OpenRouter at $0 prompt and completion pricing and already passes isFreeModel through the stealth/*-alpha rule.
    • Weight 1, the same as the other models. Each of the 5 candidates gets about 1/5 of Auto Free traffic.
    • Reasoning { enabled: true, effort: 'high' }, the same as the other models. OpenRouter lists high in this model's supported_efforts.

At request time, getAutoFreeCandidates still includes only non-exclusive models that appear in the OpenRouter models cache. If a model is delisted later, it drops out of routing automatically. This PR also removes the stale entry from the static list.

Tests

models.test.ts now asserts the updated Auto Free set: high reasoning and weight 1 for every model. The existing isFreeModel and positive-weight invariants cover the new entry.

Verification

  • Checked each Auto Free model ID against OpenRouter's live /api/v1/models response.
  • Formatting checked with oxfmt. Tests run in CI.

Remove Ling 3.0 Flash VL, which is no longer listed on OpenRouter.
@chrarnoldus chrarnoldus self-assigned this Sep 25, 2026
@kilo-code-bot

kilo-code-bot Bot commented Sep 25, 2026 •

Copy link
Copy Markdown
Contributor

Code Review Summary

Status: No Issues Found | Recommendation: Merge

Executive Summary

The incremental change switches Space Bunny Alpha's Auto Free reasoning effort from max to high in models.ts and updates the matching test title and expectation; high is a valid OpenRouterReasoningConfig effort and the assertion set stays consistent with the Auto Free model list.

Files Reviewed (2 files)
  • apps/web/src/lib/ai-gateway/models.ts
  • apps/web/src/lib/ai-gateway/models.test.ts
Previous Review Summaries (2 snapshots, latest commit b874387)

Current summary above is authoritative. Previous snapshots are kept for context only.

Previous review (commit b874387)

Status: No Issues Found | Recommendation: Merge

Executive Summary

The incremental change reduces Space Bunny Alpha's Auto Free weight from 8 to 1 with a matching test update; the new model satisfies isFreeModel, its max reasoning config is intentional per the PR description, and no dangling references remain after removing Ling 3.0 Flash VL.

Files Reviewed (2 files)
  • apps/web/src/lib/ai-gateway/models.ts
  • apps/web/src/lib/ai-gateway/models.test.ts

Previous review (commit 8faa46d)

Status: No Issues Found | Recommendation: Merge

Executive Summary

Reviewed the two-file Auto Free model routing update in apps/web/src/lib/ai-gateway/; the weight math (8/12 = 2/3), reasoning config, and removal of the delisted model are consistent with the codebase's existing free-model detection and routing logic, with no high-confidence defects found.

Files Reviewed (2 files)
  • apps/web/src/lib/ai-gateway/models.ts
  • apps/web/src/lib/ai-gateway/models.test.ts

Reviewed by deepseek-v4.1-flash · Input: 0 · Output: 0 · Cached: 0

Review guidance: REVIEW.md from base branch main

@chrarnoldus chrarnoldus changed the title Add Space Bunny Alpha to Auto Free Add Space Bunny Alpha to Auto Free and remove delisted Ling 3.0 Flash VL Sep 25, 2026
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.

2 participants