NEWS · MODELS · #398
Cohere releases North Mini Code — open-source 30B (3B active) agentic coding MoE model
Cohere has open-sourced North Mini Code, a mixture-of-experts agentic coding model sized at 30 billion parameters with 3 billion active, released under an Apache 2.0 license. The model — available on Hugging Face, OpenRouter, and Cohere's Model Vault in bf16, fp8 and w4a16 formats and compatible with OpenCode — is optimized for agentic developer workflows and reports a 33.4 score on the Artificial Analysis Coding Index, higher throughput (up to 2.8x) and ~30% better inter-token latency versus Devstral Small 2 in Cohere's tests.
KEY POINTS
- Cohere has open-sourced North Mini Code, a mixture-of-experts agentic coding model sized at 30 billion parameters with 3 billion active, released under an Apache 2.0 license.
- The model — available on Hugging Face, OpenRouter, and Cohere's Model Vault in bf16, fp8 and w4a16 formats and compatible with OpenCode — is optimized for agentic developer workflows and reports a 33.4 score on the Artificial Analysis Coding Index, higher throughput (up to 2.8x) and ~30% better inter-token latency versus Devstral Small 2 in Cohere's tests.
- An open-source, efficient agentic coding MoE from a major model provider lowers barriers for on‑prem and sovereign developer tooling and could shift adoption toward smaller, deployable coding agents.
WHY IT MATTERS
An open-source, efficient agentic coding MoE from a major model provider lowers barriers for on‑prem and sovereign developer tooling and could shift adoption toward smaller, deployable coding agents.
SOURCES & TIMELINE
1Small, efficient, and open-source — our first agentic coding model, built for the sovereign developer ecosystem. Today we're launching North Mini Code open-source. A mixture-of-experts (MoE) model, North Mini Code is Cohere's first agentic coding model, and the inaugural member of our next generation of powerful models. At 30B total parameters with just 3B active, North Mini Code delivers strong software development …
Outsized performance, right-sized footprint — translation built for speed and cost efficiency. Today, we're releasing North Small Translate, a mixture-of-experts machine translation model with strong performance across 50+ languages. Across WMT26 benchmarks,¹ North Small Translate achieves an 83.6 score across all languages, outperforming proprietary models like DeepL and Google Translate, as well as open-weight alte…
We're excited to announce the launch of North Automations , Cohere's latest step in enabling enterprises to close the AI ROI gap with ease, control, and governance built in from the ground up. The rapid proliferation of AI agents across the enterprise has created a new set of challenges that are widening the gap between agentic AI potential and its practical implementation. These challenges include: Narrow workstream…