feat(openrouter): add Qwen3-235B models

[Chesars]

Relevant issues

Addresses request from PR #13019 comment

Pre-Submission checklist

• I have Added testing in the tests/litellm/ directory, Adding at least 1 test is a hard requirement - N/A (pricing data only, no code changes)
• My PR passes all unit tests on make test-unit
• My PR's scope is as isolated as possible, it only solves 1 specific problem

Type

🆕 New Feature

Changes

Add OpenRouter Qwen3-235B-A22B models to model_prices_and_context_window.json:

Model	Input $/M	Output $/M	Context	Features
openrouter/qwen/qwen3-235b-a22b-2507	$0.071	$0.10	262,144	function_calling, tool_choice
openrouter/qwen/qwen3-235b-a22b-thinking-2507	$0.11	$0.60	262,144	function_calling, tool_choice, reasoning

Pricing verified from OpenRouter documentation:

• https://openrouter.ai/qwen/qwen3-235b-a22b-2507
• https://openrouter.ai/qwen/qwen3-235b-a22b-thinking-2507

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
litellm	Ready	Preview, Comment	Feb 5, 2026 3:36am

[greptile-apps]

Greptile Overview

Greptile Summary

This PR adds two new Qwen3-235B OpenRouter models to the pricing configuration and includes GitHub Copilot improvements.

Changes:

• Added openrouter/qwen/qwen3-235b-a22b-2507 and openrouter/qwen/qwen3-235b-a22b-thinking-2507 models to model_prices_and_context_window.json with correct pricing (verified against OpenRouter docs)
• GitHub Copilot now auto-injects required headers (eliminates manual extra_headers requirement)
• GitHub Copilot API now preserves system prompts (removed system-to-assistant conversion)
• Deprecated disable_copilot_system_to_assistant configuration flag

Scope Note:
The PR description mentions only pricing data changes, but includes a separate commit (72e38e8) with GitHub Copilot functionality improvements. Consider updating the PR description to reflect both changes.

Confidence Score: 4/5

• This PR is safe to merge with minor documentation discrepancy
• Both changes are low-risk: the Qwen3 model additions are pure data entries with verified pricing, and the GitHub Copilot changes improve UX by auto-injecting headers and preserving system prompts. Score reduced by 1 point due to PR description not matching actual scope (mentions only pricing changes but includes GitHub Copilot improvements).
• No files require special attention - all changes are straightforward and well-implemented

Important Files Changed

Filename	Overview
model_prices_and_context_window.json	Added pricing configuration for two Qwen3-235B models with correct pricing and context window values
litellm/main.py	Added auto-injection of GitHub Copilot headers in completion function for github_copilot provider
litellm/llms/github_copilot/chat/transformation.py	Removed system-to-assistant message conversion and added Copilot header injection in validate_environment
docs/my-website/docs/providers/github_copilot.md	Updated documentation to reflect auto-injection of headers and added system prompt support example
docs/my-website/docs/proxy/config_settings.md	Marked disable_copilot_system_to_assistant setting as deprecated

Sequence Diagram

sequenceDiagram
    participant User
    participant main.py
    participant GithubCopilotConfig
    participant Authenticator
    participant CopilotAPI as GitHub Copilot API

    User->>main.py: completion(model="github_copilot/gpt-4", messages=[...])
    
    alt custom_llm_provider == "github_copilot"
        main.py->>Authenticator: get_api_key()
        Authenticator-->>main.py: copilot_api_key
        main.py->>main.py: get_copilot_default_headers(api_key)
        Note over main.py: Auto-inject headers:<br/>- editor-version<br/>- copilot-integration-id<br/>- user-agent, etc.
        main.py->>main.py: Merge with extra_headers (if provided)
    end
    
    main.py->>GithubCopilotConfig: transform_request(messages, ...)
    Note over GithubCopilotConfig: System prompts preserved<br/>(no conversion to assistant)
    GithubCopilotConfig-->>main.py: transformed_messages
    
    main.py->>GithubCopilotConfig: validate_environment()
    GithubCopilotConfig->>Authenticator: get_api_key()
    Authenticator-->>GithubCopilotConfig: copilot_api_key
    GithubCopilotConfig->>GithubCopilotConfig: get_copilot_default_headers(api_key)
    GithubCopilotConfig->>GithubCopilotConfig: Add X-Initiator header
    GithubCopilotConfig-->>main.py: validated_headers
    
    main.py->>CopilotAPI: POST /chat/completions<br/>(with auto-injected headers)
    CopilotAPI-->>main.py: response
    main.py-->>User: completion response

Loading

[greptile-apps]

_{5 files reviewed, 1 comment}

_{Edit Code Review Agent Settings | Greptile}