gemma-4-31B-it-oQ8e-mtp
This model was quantized using oQ (oMLX v0.7.1) mixed-precision quantization with Improved oQe Quantization (jundot/omlx#4385).
A mixed-precision MLX checkpoint of google/gemma-4-31B-it, the 30.7B dense model of the Gemma 4 family from Google DeepMind, with text and image input, configurable thinking modes and a 256K-token context. It includes the language model, the vision encoder and a multi-token prediction (MTP) head merged from google/gemma-4-31B-it-assistant.
The effective average is 8.746 bits/weight for the complete checkpoint.
Quantization and Bit Distribution
oQ8e allocates bits by measured layer sensitivity, then rounds each group with an importance matrix (128 samples x 512 tokens) and a weighted least-squares refit of its scale and bias.
| Storage format | Logical weights | Share of total | Tensor storage | Effective bits/weight |
|---|---|---|---|---|
| Affine 8-bit, group 64 | 30.702B | 96.722% | 30.381 GiB | 8.50 |
| BF16 | 1.040B | 3.278% | 1.938 GiB | 16.00 |
| Total | 31.743B | 100% | 32.319 GiB | 8.746 |
Storage Breakdown
| Component | Logical weights | Storage |
|---|---|---|
| Language backbone | 30.697B | 32.617 GB / 30.377 GiB |
| Vision encoder | 0.576B | 1.146 GB / 1.067 GiB |
| MTP head | 0.470B | 0.939 GB / 0.875 GiB |
| Total | 31.743B | 34.702 GB / 32.319 GiB |
License
Apache-2.0, following the source model. See the Gemma 4 license.