North Mini Code 1.0
Small & efficient, designed for agentic performance and built for community input.
Cohere's open-source code model release, a 30B-parameter mixture-of-experts model with 3B active parameters, designed for agentic coding performance.
Pricing
Model Intelligence
Recent stories
Cohere released an open-source serving system that fuses the LLM decode step into one GPU kernel launch. On North Mini Code with one H100, it reports up to 1.58x vLLM performance at the tested batch size.
Cohere added MLX support, Unsloth GGUFs, oMLX work, and updated docs for North Mini Code two days after launch, with llama.cpp still under review. The broader runtime coverage makes the 30B coding model easier to run on local Mac, quantized, and self-hosted stacks.
Cohere open-sourced North Mini Code, a 30B-parameter coding MoE with 3B active parameters, 256K context, and Apache 2.0 licensing. OpenCode added it the same day, making the release immediately usable in a coding-agent client.