8-bit Open on Hugging Face
Parameters
12B
Weights (est.)
≈ 12.9 GB
Mac memory (est.)
24 GB+
Size is estimated from the parameter count and quantisation. Memory allows a fifth extra for context plus 2 GB for the system; long contexts need more.
Run it
With mlx-lm on an Apple silicon Mac:
pip install mlx-lmmlx_lm.generate --model mlx-community/Mellum2-12B-A2.5B-Thinking-8bit --prompt "Hello"
Details
- Base model
- JetBrains/Mellum2-12B-A2.5B-Thinking (quantized)
- Vendor
- JetBrains
- Task
- Text generation
- Quantisation
- 8-bit
- Licence
- apache-2.0
- Downloads
- 71 in the last 30 days · 283 all time · 0 likes
- Published
- 2026-08-11 · updated 2026-08-11 (3mo ago)
- Languages
- en
- Tags
- mellum, thinking
Other MLX builds of JetBrains/Mellum2-12B-A2.5B-Thinking
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| Mellum2-12B-A2.5B-Thinking-4bit | 4-bit | ≈ 6.8 GB | 16 GB+ | 314 |
| Mellum2-12B-A2.5B-Thinking-mxfp4 | MXFP4 | ≈ 6.8 GB | 16 GB+ | 45 |
| Mellum2-12B-A2.5B-Thinking-6bit | 6-bit | ≈ 9.9 GB | 16 GB+ | 34 |
| Mellum2-12B-A2.5B-Thinking-8bit (this one) | 8-bit | ≈ 12.9 GB | 24 GB+ | 71 |
| Mellum2-12B-A2.5B-Thinking-bf16 | bf16 | ≈ 24.3 GB | 32 GB+ | 44 |