DeepSeek · 763B (16B active) · Mixture of Experts
Multimodal V4.1 MoE — 552B backbone plus 196B Engram memory, 16B active and a quarter of V4 Flash's KV cache Check if your GPU or Mac can run DeepSeek V4.1 Flash locally — 426.4 GB min, 710.6 GB recommended.
2026-101024K context
Mixture of Experts
Total experts: 384
Active experts: 6
Active params: 16.0B
| Quant | Bits | VRAM | Quality | Status |
|---|---|---|---|---|
| Q2_K | 2 | 244.8 GB | low | — |
| Q3_K_M | 3 | 342.5 GB | moderate | — |
| Q4_K_M | 4 | 391.3 GB | good | — |
| Q5_K_M | 5 | 489 GB | good | — |
| Q6_K | 6 | 586.7 GB | excellent | — |
| Q8_0 | 8 | 782.2 GB | excellent | — |
| F16 | 16 | 1563.8 GB | lossless | — |
Can I run DeepSeek V4.1 Flash locally?
- Can I run DeepSeek V4.1 Flash locally?
- DeepSeek V4.1 Flash needs about 426.4 GB of memory at a minimum and 710.6 GB recommended. Open this page to grade it against your GPU or Mac, then run it with runai, Ollama or LM Studio.
- How much VRAM does DeepSeek V4.1 Flash need?
- At Q4_K_M, DeepSeek V4.1 Flash uses about 391.3 GB of VRAM. Higher quants need more memory; lower quants fit tighter cards with a quality tradeoff.