feat(openai): add GPT-6.1 Sol support

[WebMad]

Related GitHub Issue

Closes #1863

Description

Add GPT-6.1 Sol to native OpenAI and ChatGPT-subscription Codex catalogs without changing existing defaults.

• Native OpenAI: 1,050,000 context, 128,000 output, required low/medium/high/xhigh/max reasoning, medium default, image input, and verified Standard/Flex/Priority and long-context pricing. Cached input is 5% of input pricing.
• Codex: supported maximum context of 872,000 (the official client defaults to 272,000 but permits overriding up to 872,000), low default reasoning, subscription pricing, and Responses Lite routing.
• Cover unsupported/disabled reasoning fallback, supported effort selection, native tool requests, tier pricing, and Codex Fast-mode Responses Lite tool/header shaping.
• Existing models and provider defaults remain unchanged. No unrelated code-index commits are included.

Sources:

• https://developers.openai.com/api/docs/models/gpt-6.1-sol
• https://learn.chatgpt.com/docs/models#gpt-61-sol
• https://github.com/openai/codex/blob/main/codex-rs/models-manager/models.json
• https://github.com/openai/codex/blob/main/codex-rs/models-manager/src/model_info.rs

Test Procedure

Passed locally:

• Types package model catalog suite: 5 tests.
• Extension provider suites: 181 tests across GPT-6.1 Sol, native OpenAI, Codex, native usage, and Codex tool calls.
• Re-ran the 5 catalog tests and 12 GPT-6.1 Sol request tests after updating the Codex context to 872K.
• Types and extension type checks, repository-wide lint and type-check hooks.
• ESLint suppression counts unchanged; whitespace checks passed.

Tests ran on macOS with Node 24.7.0 and pnpm 10.8.1; the repository declares Node 22.23.1. Provider and catalog sources on the clean PR branch match the tested sources.

Manual verification for an account with access: select GPT-6.1 Sol in native OpenAI or Codex, run a coding task with tools, and verify reasoning controls and Codex Fast mode. Live paid API requests and near-limit context requests were not exercised.

Pre-Submission Checklist

• Issue Linked: Linked to [ENHANCEMENT] Add GPT-6.1 Sol support #1863; maintainer approval has not been verified.
• Scope: Changes are focused on the linked issue.
• Self-Review: Reviewed implementation and official model metadata.
• Testing: Added model and request tests.
• Visual Snapshot: Not applicable; no layout or visual component changes.
• Documentation Impact: Considered; source links included with model definitions.
• Contribution Guidelines: Read the contributor guidelines.

Visual Snapshots

Not applicable.

Videos (interaction / animation only)

Not applicable.

Documentation Updates

No separate user-facing documentation updates required; the existing model picker consumes the shared catalog.

Additional Notes

AI-assisted implementation. Ultra delegation, Ultrafast, subscription billing changes, and dots are outside scope.

Get in Touch

Please use this PR thread.

[github-actions]

Review status

Thanks for contributing. This comment tracks the review sequence and the next action.

Current step: Awaiting fresh human maintainer or CODEOWNER approval.

Automated review is complete for the latest commit but does not replace human approval.

Review-state labels are managed by this workflow; do not edit them manually.

[coderabbitai]

Navigate logical layers of code changes, visualize relationships, and explore their blast radius.

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository: Zoo-Code-Org/Zoo-Code/.coderabbit.yaml

Review profile: ASSERTIVE

Plan: Advanced

Run ID: ed49678b-6085-4621-ba6e-74647ae6c627

📥 Commits

Reviewing files that changed from the base of the PR and between acd2e0b and c79f556.

📒 Files selected for processing (1)

• turbo.json

Included review availability: This review used your included allowance. Your plan provides up to 4 included reviews per hour; 3 remain after this review.

📜 Recent review details

🧰 Additional context used

📓 Path-based instructions (1)

Act as an adversarial second-opinion reviewer.

⚙️ CodeRabbit configuration file

Files:

• turbo.json

🔇 Additional comments (1)

turbo.json (1)

18-18: LGTM!

---

📝 Summary

Summary by CodeRabbit

• New Features
  • Added GPT-6.1 Sol for OpenAI and OpenAI Codex, with image input, prompt caching, adjustable reasoning effort, and configurable response detail.
  • GPT-6.1 Sol supports up to 1,050,000 context tokens with OpenAI and 872,000 with OpenAI Codex, with output limits of up to 128,000 tokens.
  • OpenAI pricing varies by context length and service tier; OpenAI Codex pricing is listed as zero.
  • Existing default models remain unchanged.

Walkthrough

The OpenAI native and Codex model catalogs now include GPT-6.1 Sol. The Codex handler routes it through the existing Responses Lite path. Tests cover model settings and native and Codex request construction. The build task now depends on upstream package builds.

Changes

GPT-6.1 Sol support

Layer / File(s)	Summary
Model catalog entries packages/types/src/providers/openai.ts, packages/types/src/providers/openai-codex.ts, packages/types/src/__tests__/openai-models.test.ts	The native and Codex catalogs define GPT-6.1 Sol capabilities, limits, reasoning settings, and pricing. Tests check both entries and confirm the provider defaults remain unchanged.
Handler support and request validation src/api/providers/openai-codex.ts, src/api/providers/__tests__/openai-sol-6.1.spec.ts	The Codex handler adds GPT-6.1 Sol to the Responses Lite model set. Tests check reasoning selection and native and Codex request fields.

Build dependency ordering

Layer / File(s)	Summary
Upstream build dependencies turbo.json	The build task now depends on upstream package build tasks.

Priority: ➖ Normal

Estimated code review effort: 3 (Moderate) | ~20 minutes

Change: Feature · Severity of issue fixed: Medium

Sequence Diagram(s)

sequenceDiagram
  participant ZooCode as Zoo Code
  participant CodexHandler as Codex handler
  participant ResponsesClient as Responses client
  ZooCode->>CodexHandler: Send request for gpt-6.1-sol
  CodexHandler->>ResponsesClient: Send transformed Responses Lite request

Loading

Merge Risk: ⚪ Minimal · up to c79f5

The documented model settings and build ordering show no established failure. Whether Codex users can select an additional reasoning level remains unclear, but no concrete merge-blocking risk is established.

Architecture Summary

Architecture risk: 🔵 Low · up to c79f5

The change affects 3 systems.

Changed systems: packages/types, src, turbo.json

Architecture concerns
No architecture-level concerns identified.

Review details

Systems and components

• observed — packages/types (library) was modified; 3 changed files map to changed impact.
• observed — src (api) was modified; 2 changed files map to changed impact.
• observed — turbo.json (service) was modified; 1 changed file maps to changed impact.

Before / after behavior

• observed — Modified behavior in packages/types/src/tests/openai-models.test.ts: Added a GPT-6.1 Sol test that checks both default model IDs, native model capabilities and pricing (including long-context multipliers and flex/priority tiers), and Codex capabilities, context window, reasoning settings, and zero prices.
• observed — Modified behavior in packages/types/src/providers/openai-codex.ts: Added gpt-6.1-sol to openAiCodexModels with its token and context limits, supported tools and capabilities, reasoning-effort options and default, zero prices, and description.
• observed — Modified behavior in packages/types/src/providers/openai.ts: Added the gpt-6.1-sol model entry to openAiNativeModels, including its capabilities, reasoning settings, pricing, long-context multipliers, flex and priority tier prices, and description.
• observed — Modified behavior in src/api/providers/tests/openai-sol-6.1.spec.ts: Adds a telemetry mock and imports the API options, OAuth manager, stream utilities, and native and Codex handlers used by the tests.

---

Important

Pre-merge checks failed

Please resolve all errors before merging. Addressing warnings is optional.

❌ Failed checks (1 warning, 1 inconclusive)

Check name	Status	Explanation	Resolution
Regression Evidence	⚠️ Warning	The Codex catalog adds supportsVerbosity: true for gpt-6.1-sol, but packages/types/src/__tests__/openai-models.test.ts does not assert this field for the new Codex entry. This flag controls the …	Add supportsVerbosity: true to the gpt-6.1-sol Codex expectations in packages/types/src/__tests__/openai-models.test.ts. Add a focused settings component assertion only if the existing generic verbosity tests do not cover the selected…
Linked Issues check	❓ Inconclusive	The implementation satisfies the reviewable coding requirements in #1863: both catalogs add gpt-6.1-sol, native and Codex handlers route the model, tool and reasoning request fields are tested, limi…	Provide reviewable authenticated end-to-end evidence that a user with model access can complete a coding task with Zoo Code tools. Include both supported OpenAI connection paths if both are intended to be verified.

✅ Passed checks (6 passed)

Check name	Status	Explanation
Out of Scope Changes check	✅ Passed	The changes remain within #1863. The catalog entries, request routing, capability tests, and model-specific build dependency support GPT-6.1 Sol selection and use. The changes do not alter plans, bill…
Security Boundaries	✅ Passed	No changed path meets the security failure conditions. The production changes add static GPT-6.1 Sol metadata in the two model catalogs and add that fixed ID to the existing Responses Lite allowlist i…
Persistence Integrity	✅ Passed	No changed persistence path exists. The PR adds static model metadata, a Codex Responses Lite model-set entry, tests, and a Turbo build dependency. It adds no persistence write, storage update, rollba…
Lifecycle Resource Cleanup	✅ Passed	No changed lifecycle path introduces a resource leak or duplicate work. The production diff adds model metadata, adds gpt-6.1-sol to the existing RESPONSES_LITE_MODEL_IDS routing set, and adds a T…
Title check	✅ Passed	The title clearly and concisely identifies the primary change: adding GPT-6.1 Sol support for OpenAI.
Description check	✅ Passed	The description includes the linked issue, implementation details, test procedure, checklist, documentation impact, and additional notes. It clearly explains scope and testing, including untested live…

Full details: Linked Issues check

Explanation

The implementation satisfies the reviewable coding requirements in #1863: both catalogs add gpt-6.1-sol, native and Codex handlers route the model, tool and reasoning request fields are tested, limits and pricing are declared, and existing defaults remain gpt-5.6-sol. The Codex model uses Responses Lite. The available evidence does not establish an authenticated end-to-end coding task with Zoo Code tools because live paid API requests were not exercised.

Full details: Regression Evidence

Explanation

The Codex catalog adds supportsVerbosity: true for gpt-6.1-sol, but packages/types/src/__tests__/openai-models.test.ts does not assert this field for the new Codex entry. This flag controls the visible verbosity setting in webview-ui/src/components/settings/ApiOptions.tsx. A regression that removes the flag would pass the new catalog test and hide the control.

Resolution

Add supportsVerbosity: true to the gpt-6.1-sol Codex expectations in packages/types/src/__tests__/openai-models.test.ts. Add a focused settings component assertion only if the existing generic verbosity tests do not cover the selected Codex model path.

• Fix all pre-merge checks with AI

✨ Finishing Touches 💡 1

🛠️ Fix failing CI checks 💡

• Commit to this branch
• Create a new PR

🧪 Generate unit tests (beta)

• Create a new PR

---

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.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[WebMad]

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[WebMad]

@coderabbitai review

[coderabbitai]

⚠️ Action not completed

Already reviewed the last commit. Use @coderabbitai full review to rerun a review of the entire changeset.

Note: CodeRabbit is an incremental review system and does not re-review already reviewed commits. This command is applicable only when automatic reviews are paused.

[edelauna]

Thanks for adding this!

I added a small tweak to turbo.json since while testing I noticed the cached wasn't always being rebuilt when dependencies changes.