AI BriefingLLM API Conversion Library llm-rosetta Released
Key point
It's a Python-based library using a hub-and-spoke approach that can convert between various LLM API formats.
Details
llm-rosetta is a Python library that converts between different LLM API formats such as OpenAI, Anthropic, and Google GenAI.
Unlike existing LiteLLM, which provides one-way conversion, it adopts a hub-and-spoke intermediate representation (IR) to support free any-to-any conversion between all APIs. This means only one converter needs to be implemented per provider, resulting in high management efficiency.
Key features include the following:
- Zero dependencies: Works without separate runtime dependencies, designed to be lightweight using
zerodep. - Proxy functionality: Beyond being a library, it can be used as a gateway with a management panel and dashboard.
- Standards and compatibility: Passed Open Responses specification tests, and is compatible with Ollama, vLLM, LM Studio, and others.
- Shim layer: Includes a dedicated layer that handles the quirks of various providers such as OpenRouter, DeepSeek, and Qwen.
This tool has been validated for technical soundness through 5 months of operation on an actual multi-model access platform.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Jevper Released: A Jev Classification Wrapper for OpenAI-Compatible Models with Automatic Logprobs DetectionToken-Space Font Compiler Generates Fonts with Fixed Token Widths