DSPy crew ships lm15, a zero-dependency litellm alternative for multi-provider LLM calls

lateinteraction · x · 2026-09-19

The DSPy team released lm15, a lightweight alternative to litellm for calling LLMs. It works with all Foundry-hosted models (Azure OpenAI, Kimi, Anthropic), imports much faster, and has zero dependencies. Developer pamelafox has already tested it and published runnable example code for switching between OpenAI and Claude endpoints using Azure bearer tokens.

Original post →

More from coding & agent

coding & agent channel →