6-bit Open on Hugging Face
Parameters
28B
Weights (est.)
≈ 23.1 GB
Mac memory (est.)
32 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/medgemma-27b-text-it-6bit --prompt "Hello"
Details
- Base model
- google/medgemma-27b-text-it (quantized)
- Vendor
- Google · Gemma
- Task
- Text generation
- Quantisation
- 6-bit
- Licence
- health-ai-developer-foundations
- Downloads
- 63 in the last 30 days · 750 all time · 0 likes
- Published
- 2025-05-26 · updated 2025-06-09 (1y ago)
- Tags
- gemma3_text, medical, clinical-reasoning, thinking
Other MLX builds of google/medgemma-27b-text-it
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| medgemma-27b-text-it-4bit | 4-bit | ≈ 16.0 GB | 24 GB+ | 246 |
| medgemma-27b-text-it-6bit (this one) | 6-bit | ≈ 23.1 GB | 32 GB+ | 63 |
| medgemma-27b-text-it-8bit | 8-bit | ≈ 30.2 GB | 48 GB+ | 154 |
| medgemma-27b-text-it-bf16 | bf16 | ≈ 56.8 GB | 96 GB+ | 173 |