adding together ai models to litellm models json

[FelipeRodriguesGare]

Relevant issues

N/A

Pre-Submission checklist

Please complete all items before asking a LiteLLM maintainer to review your PR

• I have Added testing in the tests/litellm/ directory, Adding at least 1 test is a hard requirement - see details
• 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

CI (LiteLLM team)

CI status guideline:

• 50-55 passing tests: main is stable with minor issues.
• 45-49 passing tests: acceptable but needs attention
• <= 40 passing tests: unstable; be careful with your merges and assess the risk.

• Branch creation CI run
  Link:

• CI run for the last commit
  Link:

• Merge / cherry-pick CI run
  Links:

Type

🐛 Bug Fix

Changes

This fixes function calling in kimi-k2.5 and glm-4.7 for together_ai provider specially in claude code.

[vercel]

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

Project	Deployment	Actions	Updated (UTC)
litellm	Ready	Preview, Comment	Feb 3, 2026 0:58am

[greptile-apps]

Greptile Overview

Greptile Summary

This PR adds two new Together AI model configurations to model_prices_and_context_window.json:

• together_ai/zai-org/GLM-4.7: 200K context window, function calling and reasoning support
• together_ai/moonshotai/Kimi-K2.5: 256K context window, function calling, vision, and reasoning support

The PR description states this fixes function calling in kimi-k2.5 and glm-4.7 for the together_ai provider, specifically in Claude Code.

Key observations:

• Both models follow existing patterns in the JSON file
• Pricing and context window specifications appear consistent with similar models
• GLM-4.7 includes supports_parallel_function_calling like its sibling GLM-4.6, but Kimi-K2.5 does not include this flag despite similar models having it

Missing requirements:

• PR checklist item "Adding at least 1 test" is not checked - no tests were added in tests/litellm/
• PR checklist item "My PR passes all unit tests on make test-unit" is not checked

Confidence Score: 4/5

• This PR is safe to merge with minor risk - it only adds model configuration entries
• The changes are limited to JSON configuration entries following established patterns. The main concerns are: (1) missing test coverage as required by PR checklist, and (2) potential inconsistency in supports_parallel_function_calling flag. However, the changes are non-breaking and low-risk since they only add new model configurations without modifying existing ones.
• No files require special attention - the single JSON configuration file is straightforward

Important Files Changed

Filename	Overview
model_prices_and_context_window.json	Added two Together AI model configurations: zai-org/GLM-4.7 and moonshotai/Kimi-K2.5 with function calling support

Sequence Diagram

sequenceDiagram
    participant User as User/Application
    participant LiteLLM as LiteLLM Core
    participant Config as model_prices_and_context_window.json
    participant TogetherAI as Together AI Provider
    participant API as Together AI API

    User->>LiteLLM: completion(model="together_ai/moonshotai/Kimi-K2.5")
    LiteLLM->>Config: Lookup model configuration
    Config-->>LiteLLM: Return model config (supports_function_calling=true)
    LiteLLM->>TogetherAI: Transform request with tool parameters
    TogetherAI->>API: Forward chat completion with tools
    API-->>TogetherAI: Response with function calls
    TogetherAI-->>LiteLLM: Transform response to OpenAI format
    LiteLLM-->>User: Standardized completion with function calls
    
    Note over User,API: Same flow applies for GLM-4.7 model
    User->>LiteLLM: completion(model="together_ai/zai-org/GLM-4.7")
    LiteLLM->>Config: Lookup model configuration
    Config-->>LiteLLM: Return config (supports_parallel_function_calling=true)
    LiteLLM->>TogetherAI: Transform with parallel tool support
    TogetherAI->>API: Forward request
    API-->>TogetherAI: Response
    TogetherAI-->>LiteLLM: Standardized response
    LiteLLM-->>User: Completion result

Loading

[greptile-apps]

_{1 file reviewed, 1 comment}

_{Edit Code Review Agent Settings | Greptile}

[greptile-apps]

Check if Kimi-K2.5 should include supports_parallel_function_calling

The similar Kimi-K2-Instruct-0905 model includes supports_parallel_function_calling (line 27152), and the newly added GLM-4.7 also has it. Verify if this model should have it too for consistency.

Prompt To Fix With AI

This is a comment left during a code review.
Path: model_prices_and_context_window.json
Line: 27130:27143

Comment:
Check if `Kimi-K2.5` should include `supports_parallel_function_calling`

The similar `Kimi-K2-Instruct-0905` model includes `supports_parallel_function_calling` (line 27152), and the newly added `GLM-4.7` also has it. Verify if this model should have it too for consistency.

How can I resolve this? If you propose a fix, please make it concise.

[FelipeRodriguesGare]

Not sure is this parameter is supported. I followed the other kimi-k2.5 configuration from moonshotai.