Featherless.ai, a San Francisco-based serverless inference platform, said it will add MiniMax M3.1 Flash on November 4. The company said the open-source model is built for coding, agentic workflows, and computer-use tasks.

Featherless said the model has 428 billion parameters, with 23 billion active per token, and supports vision through an OpenAI-compatible API.

The company said M3.1 Flash scores 29 on Artificial Analysis’s Intelligence Index and costs US$0.49 per task to run the full evaluation.

The launch does not appear to be exclusive. OpenRouter already lists MiniMax M3.1 Flash and has an API pricing and benchmarks page for MiniMax M3.

🔗 Source: Featherless.ai

Recent Featherless.ai developments