Cohere has launched North Mini Code, its first agentic coding model and the first in its new generation of powerful models. The model, which combines a mixture-of-experts (MoE) architecture, is designed to deliver strong software development performance without requiring extensive hardware. It is freely available under an Apache 2.0 license and is optimized for code generation, agentic software engineering, and terminal tasks. Developers can access the model through Hugging Face, Cohere's Model Vault, or via the Cohere API. According to Cohere, the model is built for efficiency and speed, with a focus on minimizing the total cost of ownership. 'We're building in the open, because the future of AI should be shaped by the people running, testing, and improving it,' said Cohere. Source: cohere
North Mini Code achieves competitive scores across benchmarks against models of its size class, demonstrating strong performance in real-world software engineering tasks. It scores 33.4 on the Artificial Analysis Coding Index, a competitive position among similarly sized models. The model also shows a 2.8x higher output throughput than Devstral Small 2 under identical concurrency levels and hardware configurations. In practical terms, this translates to nearly three times the work rate, enabling faster iteration while reducing computational overhead. Additionally, North Mini Code demonstrates a 30% advantage in inter-token latency, a metric that reflects the consistency and pacing of token generation. Source: cohere
North Mini Code is Cohere's first open-source model for developers, designed to provide control and flexibility over agentic coding infrastructure. It is built for agentic workflows, including understanding and orchestrating sub-agents, mapping systems architecture, and running code reviews. Developers can deploy the model on-prem or locally, on their own terms. Cohere emphasized that community feedback will directly shape its roadmap as it expands the ecosystem toward more open and sovereign developer models. Source: cohere