You can deploy this model and start earning money today!

pipenetwork/
Gemma-4-31B-it-MLX-8bit

Name: Gemma-4-31B-it-MLX-8bit
Availability: InStock

Views

No views yet

Gemma-4-31B-it-MLX-8bit

MLX (Apple Silicon) conversion of google/gemma-4-31B-it, quantized to 8-bit · near-lossless.

Text-only build of the Gemma-4 31B backbone (the multimodal vision components are not included).

bash

pip install mlx-lm
python -m mlx_lm generate --model pipenetwork/Gemma-4-31B-it-MLX-8bit --prompt "Explain attention in transformers." -m 256

Smoke-tested locally: loads and generates coherent text.

Apache 2.0 (inherited from the base model). Quantization config: {"group_size": 64, "bits": 8, "mode": "affine"}.