DeepSeek-V4.1-Flash-1017-oQ3e-mtp
mlx-community/DeepSeek-V4.1-Flash-1017-oQ3e-mtp
Vision + text · mlx
Parameters
112B
Weights (est.)
≈ 381 GB
Mac memory (est.)
512 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-vlm on an Apple silicon Mac:
pip install mlx-vlmpython -m mlx_vlm.generate --model mlx-community/DeepSeek-V4.1-Flash-1017-oQ3e-mtp --image photo.jpg --prompt "Describe this image"
Details
- Base model
- deepseek-ai/DeepSeek-V4.1-Flash-1017 (quantized) · DeepSeek V4.1 Flash 1017 on the radar →
- Vendor
- DeepSeek
- Task
- Vision + text
- Quantisation
- Not stated
- Licence
- mit
- Downloads
- 1.6K in the last 30 days · 1.6K all time · 1 likes
- Published
- 2026-10-17 · updated 2026-10-29 (13d ago)
- Tags
- deepseek_v41, quantized
Other MLX builds of deepseek-ai/DeepSeek-V4.1-Flash-1017
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| DeepSeek-V4.1-Flash-1017-oQ2e-mtp | — | ≈ 313 GB | 512 GB+ | 2.1K |
| DeepSeek-V4.1-Flash-1017-oQ2.5e-mtp | — | ≈ 343 GB | 512 GB+ | 4.0K |
| DeepSeek-V4.1-Flash-1017-oQ3e-mtp (this one) | — | ≈ 381 GB | 512 GB+ | 1.6K |
| DeepSeek-V4.1-Flash-1017-oQ2e-ngram-mtp | — | ≈ 387 GB | 512 GB+ | 1.0K |
| DeepSeek-V4.1-Flash-1017-oQ2.5e-ngram-mtp | — | ≈ 416 GB | 512 GB+ | 1.3K |
| DeepSeek-V4.1-Flash-1017-oQ4e-mtp | — | ≈ 432 GB | 512 GB+ | 8.2K |
| DeepSeek-V4.1-Flash-1017-oQ3e-ngram-mtp | — | ≈ 455 GB | 512 GB+ | 1.8K |
| DeepSeek-V4.1-Flash-1017-oQ4e-ngram-mtp | — | ≈ 506 GB | > 512 GB | 3.0K |