North Mini Code
Cohere's first model for developers.
North Mini Code is Cohere's 30B total / 3B active parameter Mixture-of-Experts model trained for agentic coding, code generation, agentic software engineering, and terminal tasks; it is released under Apache 2.0 and available via weights, Cohere API, Model Vault, and other channels.
Pricing
Publicly described as free on Hugging Face and Model Vault; no standalone numeric API or instance price is stated on Cohere’s public pricing pages.
Cohere’s public materials do not publish a numeric token price for North Mini Code. The launch post explicitly says North Mini Code is “available for free on Hugging Face and Model Vault,” while the pricing page routes North-related enterprise usage to contact sales/custom pricing and does not list a standalone rate for North Mini Code.
Model Intelligence
Recent stories
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.