Just when you thought the open-weight coding model race couldn't get any tighter, Cohere quietly dropped North Mini Code — and it's a serious contender.
Released on June 9, 2026, North Mini Code is Cohere's first open-source agentic coding model, built on a 30B-parameter Mixture-of-Experts (MoE) architecture with just 3B active parameters. That means it runs on consumer hardware while delivering frontier-adjacent coding performance.
Here's what makes it stand out.
Sovereign by Design
Cohere has been hammering the "AI sovereignty" drum for years, and North Mini Code is the embodiment. It's designed for enterprises and developers who want to fine-tune, deploy, and own their entire stack — no cloud dependency, no API gatekeeping. You can run it on-prem, air-gapped, or in your own VPC.
Performance That Punches Above Its Weight
Despite its small active footprint, North Mini Code scores an impressive 27.6 on the Artificial Analysis Intelligence Index — competitive with models 3–5x its size. On SWE-bench (software engineering), it holds its own against much larger peers, making it a strong option for code generation, bug fixing, and agentic workflows.
Why Now?
We're 33 days out from the EU AI Act's August 2, 2026 transparency deadline, and the industry is scrambling. Open-weight models like North Mini Code give organizations a compliance-friendly path — full control over data, no opaque third-party API calls, and the ability to audit everything.
The Bottom Line
Cohere just proved you don't need a 500B monster to do real coding work. North Mini Code is lean, capable, sovereign, and open. If you're building an AI-powered dev toolchain in 2026, this needs to be on your radar.
Have you tried North Mini Code yet? Drop your thoughts below.
Top comments (0)