North Mini Code — local AI model by Cohere
Written by Jakub Rusinowski · Last updated
Cohere's first model aimed at developers rather than enterprises, and the opening entry in their North family of code agents. A 30B sparse MoE activating ~3B per token, Apache 2.0, 256K context and 64K maximum generation — small enough to run locally while posting a Coding Index of 33.4, ahead of several open models many times its size.
Variants
The smallest North Mini Code variant needs about 19 GB of VRAM at Q4_K_M — quantized weights plus framework overhead, before any KV cache.
| Model | VRAM at Q4 | VRAM | Context | Run it |
|---|---|---|---|---|
| North Mini Code 1.0 30B-A3B → 30B (3B active) | ~18.9 GB | 256,000 | ollama run north-mini-code |
Memory is quantized weights plus overhead at Q4_K_M, from the same engine as the GPU & VRAM checker.
How to run North Mini Code locally
Install Ollama, then pull the tag.
ollama run north-mini-codePick a size above for its own VRAM figure, speed estimate and install command.
Licence
Commercial use permitted. No usage restrictions beyond attribution.
Applies to: North Mini Code 1.0 30B-A3BRecommended GPU
The cheapest catalogued GPU that runs North Mini Code locally (min 19 GB VRAM) is the AMD Radeon RX 7900 XTX (24 GB).
Can I run North Mini Code on my GPU?
- North Mini Code on AMD Radeon RX 7800 XT
- North Mini Code on AMD Radeon RX 7900 XT
- North Mini Code on AMD Radeon RX 7900 XTX
- North Mini Code on AMD Radeon RX 9060 XT 16GB
- North Mini Code on AMD Radeon RX 9060 XT 8GB
- North Mini Code on AMD Radeon RX 9070
- North Mini Code on AMD Radeon RX 9070 XT
- North Mini Code on Apple M1
- North Mini Code on Apple M1 Pro
- North Mini Code on Apple M2
- North Mini Code on Apple M2 Pro
- North Mini Code on Apple M3
- North Mini Code on Apple M3 Pro
- North Mini Code on Apple M4
- North Mini Code on Apple M4 Pro
- North Mini Code on Apple M5
- North Mini Code on Intel Arc B570
- North Mini Code on Intel Arc B580
- North Mini Code on NVIDIA GeForce RTX 3060 (12GB)
- North Mini Code on NVIDIA GeForce RTX 3070
- North Mini Code on NVIDIA GeForce RTX 3070 Ti
- North Mini Code on NVIDIA GeForce RTX 3080 (10GB)
- North Mini Code on NVIDIA GeForce RTX 3090
- North Mini Code on NVIDIA GeForce RTX 4060
- North Mini Code on NVIDIA GeForce RTX 4060 Ti 16GB
- North Mini Code on NVIDIA GeForce RTX 4070
- North Mini Code on NVIDIA GeForce RTX 4070 Super
- North Mini Code on NVIDIA GeForce RTX 4070 Ti
- North Mini Code on NVIDIA GeForce RTX 4070 Ti Super
- North Mini Code on NVIDIA GeForce RTX 4080
- North Mini Code on NVIDIA GeForce RTX 4080 Super
- North Mini Code on NVIDIA GeForce RTX 4090
- North Mini Code on NVIDIA GeForce RTX 5060
- North Mini Code on NVIDIA GeForce RTX 5060 Ti 16GB
- North Mini Code on NVIDIA GeForce RTX 5060 Ti 8GB
- North Mini Code on NVIDIA GeForce RTX 5070
- North Mini Code on NVIDIA GeForce RTX 5070 Ti
- North Mini Code on NVIDIA GeForce RTX 5080
- North Mini Code on NVIDIA GeForce RTX 5090
North Mini Code — frequently asked questions
How much VRAM does North Mini Code need?
North Mini Code needs about 19 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: North Mini Code 1.0 30B-A3B (19 GB, Q4_K_M). On Apple Silicon, unified memory counts toward this requirement.
Can I run North Mini Code on an RTX 4090 (24 GB)?
Yes — North Mini Code runs on an RTX 4090 (24 GB) and other 24 GB cards such as the RTX 3090. Smaller variants also fit comfortably on 8–16 GB GPUs at Q4_K_M.
What quantization should I use for North Mini Code?
Q4_K_M is the best balance of quality and VRAM for North Mini Code in most cases. Choose Q8_0 for near-lossless quality if you have spare VRAM, or smaller quants (Q3/Q2) only when memory is tight.
How do I run North Mini Code with Ollama?
North Mini Code has no local Ollama tag — the published tag is cloud-hosted, so running it sends your prompts to a hosted GPU rather than your own machine.