DeepSeek-V4.1-Flash-1017-oQ2e-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 deepseek-ai/DeepSeek-V4.1-Flash-1017, a minor update to DeepSeek-V4.1-Flash with improved reasoning, coding, agentic and visual performance. It is a multimodal mixture-of-experts model with 552B backbone parameters, a Causal Encoder-Decoder layout, Compressed Sparse Attention 2 (CSA2), Engram conditional memory and DSpark speculative decoding. It includes the language model, the Engram tables, the vision encoder, the multimodal aligner and the DSpark MTP weights.

The effective average is 2.688 bits/weight excluding the Engram tables, or 3.284 bits/weight for the complete checkpoint.

Quantization and Bit Distribution

oQ2e allocates bits by measured layer sensitivity, then rounds each group with an importance matrix (1024 samples x 512 tokens) and a weighted least-squares refit of its scale and bias. Below oQ4, the Engram tables stay at 4 bits and are left out of the bit budget, because oMLX can serve them from SSD instead of memory.

Storage format	Logical weights	Share of total	Tensor storage	Effective bits/weight
Affine 2-bit, group 64	543.582B	71.224%	158.203 GiB	2.50
MXFP4, group 32	13.590B	1.781%	6.724 GiB	4.25
Affine 8-bit, group 64	5.200B	0.681%	5.145 GiB	8.50
MXFP8, group 32	0.779B	0.102%	0.748 GiB	8.25
BF16	3.399B	0.445%	6.332 GiB	16.00
FP32	0.042B	0.006%	0.158 GiB	32.00
Engram affine 4-bit, group 32	196.614B	25.762%	114.444 GiB	5.00
Total	763.205B	100%	291.754 GiB	3.284

Storage Breakdown

Component	Logical weights	Storage
Language backbone	551.881B	181.366 GB / 168.911 GiB
Engram tables	196.614B	122.884 GB / 114.444 GiB
Vision encoder	0.485B	0.971 GB / 0.904 GiB
MTP head	14.225B	8.047 GB / 7.495 GiB
Total	763.205B	313.268 GB / 291.754 GiB

License

MIT, following the source model. See LICENSE.