baa-ai/Gemma-4-31B-it-RAM-30GB-MLX

Gemma-4-31B-it — 30GB (MLX)

Mixed-precision quantized version of google/gemma-4-31B-it optimised by baa.ai using a proprietary Black Sheep AI method.

Per-tensor bit-width allocation via advanced sensitivity analysis with adjusted vision encoder allocation.

Metrics

Metric	Value
Size	30 GB
Average bits	7.8
MMLU vs BF16	100% of BF16

Usage

from mlx_lm import load, generate

model, tokenizer = load("baa-ai/Gemma-4-31B-it-RAM-30GB-MLX")
response = generate(model, tokenizer, prompt="Hello!", max_tokens=256)
print(response)

---

Quantized by baa.ai