DeepSeek · 1.6T (49B active) · Mixture of Experts
Flagship V4 MoE — 49B active, 1M context Check if your GPU or Mac can run DeepSeek V4 Pro locally — 894.1 GB min, 1490.1 GB recommended.
2026-041024K context
Mixture of Experts
Total experts: 384
Active experts: 6
Active params: 49.0B
| Quant | Bits | VRAM | Quality | Status |
|---|---|---|---|---|
| Q2_K | 2 | 512.7 GB | low | — |
| Q3_K_M | 3 | 717.6 GB | moderate | — |
| Q4_K_M | 4 | 820.1 GB | good | — |
| Q5_K_M | 5 | 1025 GB | good | — |
| Q6_K | 6 | 1229.8 GB | excellent | — |
| Q8_0 | 8 | 1639.6 GB | excellent | — |
| F16 | 16 | 3278.8 GB | lossless | — |
Can I run DeepSeek V4 Pro locally?
- Can I run DeepSeek V4 Pro locally?
- DeepSeek V4 Pro needs about 894.1 GB of memory at a minimum and 1490.1 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 Pro need?
- At Q4_K_M, DeepSeek V4 Pro uses about 820.1 GB of VRAM. Higher quants need more memory; lower quants fit tighter cards with a quality tradeoff.