Cohere Launches North Mini Code, an Open-Source Agentic Coding Model
Key Takeaways
- ▸North Mini Code is Cohere's inaugural open-source agentic coding model, combining 30B total parameters with only 3B active parameters for efficient deployment without demanding extensive hardware
- ▸The model delivers 2.8x higher output throughput and 30% better inter-token latency compared to similarly-sized competitors, enabling faster iteration with lower computational overhead
- ▸Released under Apache 2.0 license with free availability on Hugging Face and Model Vault, supporting sovereign AI infrastructure for developers seeking vendor independence and on-premise deployment
Summary
Cohere has released North Mini Code, an open-source mixture-of-experts (MoE) coding model with 30 billion total parameters and 3 billion active parameters, available under an Apache 2.0 license. The model is designed as Cohere's first agentic coding model, enabling developers to run sophisticated coding agents locally or on-premises without extensive hardware requirements. North Mini Code delivers competitive performance across coding benchmarks, achieving a 33.4 score on the Artificial Analysis Coding Index while demonstrating significant efficiency gains, including 2.8x higher output throughput compared to competitors like Devstral Small 2 and 30% better inter-token latency. Available for free on Hugging Face, Cohere's Model Vault inference platform, and compatible with OpenCode and most coding agents, the model is specifically optimized for agentic workflows including sub-agent orchestration, system architecture mapping, and automated code reviews.
- Specifically engineered for agentic capabilities including autonomous sub-agent orchestration, system architecture analysis, and code review automation
Editorial Opinion
North Mini Code represents a significant move by Cohere toward democratizing agentic AI for developers. By releasing a capable coding model under a permissive open-source license with strong efficiency metrics, Cohere directly addresses developer demand for sovereign alternatives to proprietary cloud APIs. This launch signals a broader industry shift where open-source models aren't just alternatives—they're becoming the preferred choice for developers prioritizing control, transparency, and cost-effectiveness.
