feat(usage): register Cognition/Devin list prices for both devin surfaces - #4466
Conversation
…aces The devin and devin-cli providers had no rows in the generated metadata bundle or the expected-price overlays, so every attempt resolved price_unmatched. Cognition publishes per-model list rates in the official Devin docs modelCostData table (docs.devin.ai/desktop/models, verified 2026-09-13); both self-serve overage and enterprise ACU conversion bill at those rates. Register them as verified-derived estimates on the subscription/ACU surface, covering the full liveModels roster for both provider ids. Time-boxed promos (SWE-2 free self-serve through 2026-10-08, 75% off enterprise through 2026-12-31) are noted in the source strings but not baked into the tuples. gemini-3-8-flash is absent from the Devin table and derives from Google's published rate instead.
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. |
|
✅ Deterministic PR hygiene checks passed. |
|
Caution Review failedThe pull request is closed. ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: ASSERTIVE Plan: Advanced Run ID: 📒 Files selected for processing (2)
📝 WalkthroughWalkthroughThe change adds verified Cognition/Devin pricing data for ChangesDevin pricing support
Estimated code review effort: 2 (Simple) | ~12 minutes Change: Feature ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
리뷰 · 우선순위 66 / 80이 PR은 Cognition/Devin 쪽 사용량(Usage) 추정에서 계속 구체적으로 docs.devin.ai/desktop/models 의 공식 레지스트리 정적 seed와 비교하면 의도가 분명하다. 전체적으로 “Devin을 쓰기 시작한 사람이 Usage에서 가격을 못 보는” 실제 빈칸을, 이미 검증된 overlay 계약으로 채우는 PR이다. 우선순위 66은 그 실용성(높음)과 위험(낮음, 데이터만)의 균형이다. 머지 전에 docs 표 숫자 몇 개만 눈으로 맞추면 충분하다.
메인테이너의 판단이 필요한 지점
너의 추천 이 댓글은 grok-bot이 작성했습니다 |
![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: 38c1e245e1
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| // gpt-5-6-sol uses the enterprise list column — the same table's self-serve | ||
| // column shows a discounted 1.2/6, and the doc calls the list rate the | ||
| // billing rate for overage. glm-5-2 likewise takes the nonzero list column. | ||
| { provider: "devin-cli", modelId: "swe-2", cost4: DEVIN_SWE_2, source: `${DEVIN_SWE2_NOTE}; ${DEVIN_PRICING}`, verifiedAt: "2026-09-13", status: "verified-derived" }, |
There was a problem hiding this comment.
This changes pricing behavior under src/usage/, but the source-to-doc map assigns that area to both structure/runtime.md and structure/gui-and-management-api.md, and neither is updated; the latter's expected-price-overlay contract currently discusses only the Z.AI derived rows. Update both owning documents in this commit to reflect the new Devin overlays and their subscription/list-rate semantics.
AGENTS.md reference: src/AGENTS.md:L11-L11
Useful? React with 👍 / 👎.
| // Cognition/Devin — the two OAuth surfaces resolve by exact provider id, so | ||
| // each carries the roster its liveModels discovery can surface. swe-2 and | ||
| // swe-1-6 are listed on both even though each static seed names only one | ||
| // side: the live catalog is authoritative and drifts between them. |
There was a problem hiding this comment.
The two lists are still based mostly on their static seeds even though fetchDevinUsableModels applies no provider-specific filtering and both registry entries share the union-like DEVIN_MODEL_CONTEXT_WINDOWS. If a devin-cli account discovers gpt-5-6-luna/gpt-5-6-terra, or a devin account discovers gpt-6-astra, the exact overlay lookup misses; the dashed GPT IDs do not match the generated dotted metadata IDs, and Astra has no generated vendor row, so those attempts remain price_unmatched. Register the full discoverable base-model union for both provider namespaces, with tests for the cross-surface cases.
Useful? React with 👍 / 👎.
1 participant
Summary
Verification
Checklist
Summary by CodeRabbit
New Features
Tests