Skip to content
AI Model Radar

North Mini Code 1.0

CohereLabs/North-Mini-Code-1.0
Open Source · chat · open-weights
GA Alert me on changes
Context
256K
Max output
64K
Input $/1M
$0.00
Output $/1M
$0.00
Modalities
text
Released
05 Jun 2026
License: apache-2.0 · CohereLabs/North-Mini-Code-1.0
AI summary
● machine-written

Cohere releases North Mini Code 1.0, sparse MoE coding model

North Mini Code 1.0 is Cohere's open-weight sparse mixture-of-experts model with 30B total parameters and 3B active parameters, optimized for code generation, agentic software engineering, and terminal tasks. The model features a 256K-token context window, supports tool use and reasoning, and is available under the Apache 2.0 license for commercial use. It enables low-latency inference on local hardware and is currently available free through multiple providers.

What's new
  • Sparse MoE architecture with 30B total / 3B active parameters for efficient inference
  • 256K-token context window with up to 64K-token maximum output
  • Interleaved reasoning and tool use via JSON schema
  • Open-weight release under Apache 2.0 license
  • Optimized for agent harnesses like OpenCode and SWE-Agent
Best for
Code generation and software engineering tasksAgentic AI workflows and terminal automationLocal deployment on single GPUsTool-use and function-calling applications
Sources

Source: https://huggingface.co/CohereLabs/North-Mini-Code-1.0