4-bit Open on Hugging Face
Parameters
15B
Weights (est.)
≈ 8.3 GB
Mac memory (est.)
16 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/Hermes-4-14B-4bit --prompt "Hello"
Details
- Base model
- NousResearch/Hermes-4-14B (quantized)
- Vendor
- NousResearch
- Task
- Text generation
- Quantisation
- 4-bit
- Licence
- apache-2.0
- Downloads
- 1.2K in the last 30 days · 10K all time · 6 likes
- Published
- 2025-09-03 · updated 2025-09-03 (1y ago)
- Languages
- en
- Tags
- qwen3, Qwen-3-14B, instruct, finetune, reasoning, hybrid-mode, chatml, function calling, tool use, json mode, structured outputs, atropos
Other MLX builds of NousResearch/Hermes-4-14B
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| Hermes-4-14B-4bit (this one) | 4-bit | ≈ 8.3 GB | 16 GB+ | 1.2K |
| Hermes-4-14B-5bit | 5-bit | ≈ 10.2 GB | 16 GB+ | 63 |
| Hermes-4-14B-6bit | 6-bit | ≈ 12.0 GB | 16 GB+ | 86 |
| Hermes-4-14B-8bit | 8-bit | ≈ 15.7 GB | 24 GB+ | 139 |
| NousResearch_Hermes-4-14B-BF16-abliterated-mlx | bf16 | ≈ 29.5 GB | 48 GB+ | 140 |