Introducing North Mini Code: Cohere’s First Model For Developers

Cohere has released North Mini Code, a 30B-parameter Mixture-of-Experts model with 3B active parameters designed for agentic software engineering tasks. Available on Hugging Face under an Apache 2.0 license, the model demonstrates strong performance on complex code generation and agentic coding benchmarks, outperforming other models in its size class. Its architecture is a decoder-only Transformer with interleaved sliding-window and global attention, using an MoE block with 128 experts. The model underwent a multi-stage post-training process involving supervised fine-tuning (SFT) and reinforcement learning with verifiable rewards (RLVR) to enhance its coding capabilities and ensure robustness across various agent harnesses.