Hugging FaceA 30B, 3.3B active chat model from Cohere.
- Rating
- 2.9
- Downloads
- 258K
- Likes
- 153
- Parameters
- 30B, 3.3B active
- Context
- 488K
- Licence
- apache-2.0
Runs well on M4 · 16 GB
The UD-IQ2_M size from unsloth is the one to get: a 9.9 GB download that uses 11 GB of memory, and writes about 18 words a second.
Ratings
2.9 out of 5, across the tests below.
SWE-bench Verified: real GitHub issues fixed so the project's tests pass. Terminal-Bench 2.0: jobs finished alone in a terminal. Scores are the ones filed on the model's Hugging Face page or the board's own table, read 2026-11-05; some come from the maker.
North Mini Code 1.0 runs well from a Mac with an M1 and 16 GB of memory. 56 of 59 Mac configurations run it well.
Compare with
The nearest models in size.
- Qwen3 30B A3B 2507Alibaba Qwen · 660K downloadsRuns well · 9.7 GB · 9.9 words/s
- Qwen3 30B A3BAlibaba Qwen · 232K downloads · Arena 1327Runs well · 9.7 GB · 9.9 words/s
- Gemma 4 31BGoogle · 749K downloads · Arena 1452Runs well · 8.5 GB · 5.7 words/s
- Artemis 31B V1.2TheDrummer · 92K downloadsRuns well · 9.6 GB · 5.3 words/s
Sizes
From unsloth/North-Mini-Code-1.0-GGUF on Hugging Face. A smaller size is a smaller download and a little less accurate.
| Size | Download | Memory at 8K | Get |
|---|---|---|---|
| UD-IQ1_M | 9.4 GB | 9.8 GBestimated | Open |
| UD-IQ2_XXS | 9.8 GB | 10 GBestimated | Open |
| UD-IQ2_M | 9.9 GB | 10 GBestimated | Open |
| UD-Q2_K_XL | 10 GB | 11 GBestimated | Open |
| UD-IQ3_XXS | 12 GB | 12 GBestimated | Open |
| UD-IQ3_S | 13 GB | 13 GBestimated | Open |
| UD-Q3_K_M | 14 GB | 15 GBestimated | Open |
| UD-Q3_K_XL | 14 GB | 15 GBestimated | Open |
| UD-IQ4_XS | 15 GB | 16 GBestimated | Open |
| UD-IQ4_NL | 16 GB | 16 GBestimated | Open |
| UD-Q4_K_S | 18 GB | 18 GBestimated | Open |
| MXFP4_MOE | 19 GB | 19 GBestimated | Open |
| UD-Q4_K_M | 19 GB | 20 GBestimated | Open |
| UD-Q4_K_XL | 19 GB | 20 GBestimated | Open |
| UD-Q5_K_S | 22 GB | 22 GBestimated | Open |
| UD-Q5_K_M | 23 GB | 23 GBestimated | Open |
| UD-Q5_K_XL | 23 GB | 23 GBestimated | Open |
| UD-Q6_K | 26 GB | 26 GBestimated | Open |
| UD-Q6_K_XL | 28 GB | 28 GBestimated | Open |
| Q8_0 | 32 GB | 33 GBestimated | Open |
| UD-Q8_K_XL | 33 GB | 34 GBestimated | Open |
| BF16 | 61 GB | 61 GBestimated | Open |
Run It Anywhere
North Mini Code 1.0 is an open model, so any app that runs GGUF files can run it. These take the UD-IQ2_M size from unsloth straight from Hugging Face.
- LM StudioMac, Windows, Linux
Open - JanMac, Windows, Linux
Open - MstyMac, Windows, Linux
Open - OllamaMac, Windows, Linux
ollama run hf.co/unsloth/North-Mini-Code-1.0-GGUF:UD-IQ2_M - llama.cppMac, Windows, Linux
llama-server -hf unsloth/North-Mini-Code-1.0-GGUF:UD-IQ2_M - Docker Model RunnerMac, Windows, Linux
docker model run hf.co/unsloth/North-Mini-Code-1.0-GGUF:UD-IQ2_M
On a PC Graphics Card
| Video memory | Largest size that fits | Download | Memory at 8K |
|---|---|---|---|
| 6 GB | None worth running | needs 10 GB | |
| 8 GB | None worth running | needs 10 GB | |
| 10 GB | None worth running | needs 10 GB | |
| 11 GB | UD-Q2_K_XL | 10 GB | 11 GB |
| 12 GB | UD-Q2_K_XL | 10 GB | 11 GB |
| 16 GB | UD-IQ4_XS | 15 GB | 16 GB |
| 20 GB | UD-Q4_K_XL | 19 GB | 20 GB |
| 24 GB | UD-Q5_K_XL | 23 GB | 23 GB |
| 32 GB | UD-Q6_K_XL | 28 GB | 28 GB |
| 48 GB | UD-Q8_K_XL | 33 GB | 34 GB |
The largest size, of those that keep three bits a weight or more, whose memory at an 8K chat is under 92% of the card's; the rest is the desktop's. Estimated from the files, the same way as for a Mac; speed on a PC is not estimated yet. A size that doesn't fit the card still runs, slower, with part of it in ordinary memory.
On every Mac
| Mac | Size | Fit | Memory | Writes |
|---|---|---|---|---|
| M6 · 16 GB | UD-IQ2_M | Runs well | 11 GB | 30 |
| M6 · 24 GB | UD-IQ4_NL | Runs well | 16 GB | 23 |
| M6 · 32 GB | Q3_K_XL | Runs great | 15 GB | 24 |
| M5 · 16 GB | UD-IQ2_M | Runs well | 11 GB | 28 |
| M5 · 24 GB | UD-IQ4_NL | Runs well | 16 GB | 22 |
| M5 · 32 GB | Q3_K_XL | Runs great | 15 GB | 23 |
| M5 Pro · 24 GB | UD-IQ4_NL | Runs well | 16 GB | 40 |
| M5 Pro · 48 GB | UD-Q5_K_S | Runs great | 23 GB | 33 |
| M5 Pro · 64 GB | UD-Q6_K_XL | Runs great | 30 GB | 28 |
| M5 Max · 36 GB | IQ4_XS | Runs great | 17 GB | 54 |
| M5 Max · 48 GB | UD-Q5_K_S | Runs great | 23 GB | 58 |
| M5 Max · 64 GB | UD-Q6_K_XL | Runs great | 30 GB | 50 |
| M5 Max · 128 GB | Q8_0 | Runs great | 37 GB | 54 |
| M5 Ultra · 96 GB | Q8_0 | Runs great | 37 GB | 89 |
| M5 Ultra · 256 GB | Q8_0 | Runs great | 37 GB | 89 |
| M5 Ultra · 512 GB | Q8_0 | Runs great | 37 GB | 89 |
| M4 · 16 GB | UD-IQ2_M | Runs well | 11 GB | 18 |
| M4 · 24 GB | UD-IQ4_NL | Runs well | 16 GB | 15 |
| M4 · 32 GB | Q3_K_XL | Runs great | 15 GB | 15 |
| M4 Pro · 24 GB | UD-IQ4_NL | Runs well | 16 GB | 26 |
| M4 Pro · 48 GB | UD-Q5_K_S | Runs great | 23 GB | 22 |
| M4 Pro · 64 GB | UD-Q6_K_XL | Runs great | 30 GB | 20 |
| M4 Max · 36 GB | IQ4_XS | Runs great | 17 GB | 31 |
| M4 Max · 48 GB | UD-Q5_K_S | Runs great | 23 GB | 33 |
| M4 Max · 64 GB | UD-Q6_K_XL | Runs great | 30 GB | 30 |
| M4 Max · 96 GB | Q8_0 | Runs great | 37 GB | 27 |
| M4 Max · 128 GB | Q8_0 | Runs great | 37 GB | 32 |
| M3 · 8 GB | IQ2_XXS | Too big | 13 GB | |
| M3 · 16 GB | UD-IQ2_M | Runs well | 11 GB | 17 |
| M3 · 24 GB | UD-IQ4_NL | Runs well | 16 GB | 13 |
| M3 Pro · 18 GB | Q2_K_L | Runs well | 12 GB | 21 |
| M3 Pro · 36 GB | IQ4_XS | Runs great | 17 GB | 18 |
| M3 Max · 36 GB | IQ4_XS | Runs great | 17 GB | 28 |
| M3 Max · 48 GB | UD-Q5_K_S | Runs great | 23 GB | 30 |
| M3 Max · 64 GB | UD-Q6_K_XL | Runs great | 30 GB | 27 |
| M3 Max · 96 GB | Q8_0 | Runs great | 37 GB | 23 |
| M3 Max · 128 GB | Q8_0 | Runs great | 37 GB | 28 |
| M3 Ultra · 96 GB | Q8_0 | Runs great | 37 GB | 42 |
| M3 Ultra · 256 GB | Q8_0 | Runs great | 37 GB | 42 |
| M3 Ultra · 512 GB | Q8_0 | Runs great | 37 GB | 42 |
| M2 · 8 GB | IQ2_XXS | Too big | 13 GB | |
| M2 · 16 GB | UD-IQ2_M | Runs well | 11 GB | 17 |
| M2 · 24 GB | UD-IQ4_NL | Runs well | 16 GB | 13 |
| M2 Pro · 16 GB | UD-IQ2_M | Runs well | 11 GB | 27 |
| M2 Pro · 32 GB | Q3_K_XL | Runs great | 15 GB | 23 |
| M2 Max · 32 GB | Q3_K_XL | Runs great | 15 GB | 34 |
| M2 Max · 64 GB | UD-Q6_K_XL | Runs great | 30 GB | 27 |
| M2 Max · 96 GB | Q8_0 | Runs great | 37 GB | 28 |
| M2 Ultra · 64 GB | UD-Q6_K_XL | Runs great | 30 GB | 39 |
| M2 Ultra · 128 GB | Q8_0 | Runs great | 37 GB | 41 |
| M2 Ultra · 192 GB | Q8_0 | Runs great | 37 GB | 41 |
| M1 · 8 GB | IQ2_XXS | Too big | 13 GB | |
| M1 · 16 GB | UD-IQ2_M | Runs well | 11 GB | 11 |
| M1 Pro · 16 GB | UD-IQ2_M | Runs well | 11 GB | 21 |
| M1 Pro · 32 GB | Q3_K_XL | Runs great | 15 GB | 19 |
| M1 Max · 32 GB | Q3_K_XL | Runs great | 15 GB | 26 |
| M1 Max · 64 GB | UD-Q6_K_XL | Runs great | 30 GB | 21 |
| M1 Ultra · 64 GB | UD-Q6_K_XL | Runs great | 30 GB | 28 |
| M1 Ultra · 128 GB | Q8_0 | Runs great | 37 GB | 31 |
Writes is words a second, estimated from the Mac's memory speed and the model's size; how, and how wrong.
Packaged by
2 repositories on Hugging Face carry this model for llama.cpp. Outmake reads them all as one model.
Information
- Maker
- Cohere
- Released
- June 2026
- Parameters
- 30B, 3.3B active
- Longest chat
- 488K tokens
- Reads pictures
- No
- Licence
- apache-2.0
- Downloads this month
- 258,300
- Likes
- 153
- Sizes
- 22, 9.4 to 61 GB
- Kind
- Maker's release
Badge for the model card
Shows on North Mini Code 1.0's Hugging Face card which Macs run it well, linked to this page. It updates every morning.
Files, downloads and likes read from Hugging Face on November 5, 2026 and refreshed every morning. Memory and speed are estimates until someone running Outmake measures this model on that Mac.