Can I run DeepSeek V4 Flash locally?
DeepSeek V4 Flash by deepseek-ai has 291B parameters (14B active per token, 256 experts, 6 used per token). At 4-bit with an 8K context it needs about 167 GB of GPU memory. It fits on 6 of 65 GPUs and Macs we track.
This model uses a newer compressed or sparse attention design. Our KV-cache numbers use the standard formula, so treat long-context figures as an upper bound.
Memory needed by quant and context length
GB of GPU memory = weights + KV cache (FP16) + runtime overhead. Weights use parameter count × bits per weight. How we calculate.
| Quant | Quality | 4K ctx | 8K ctx | 32K ctx | 128K ctx |
|---|---|---|---|---|---|
| 2-bit (Q2_K) | rough | 103 GB | 103 GB | 105 GB | 114 GB |
| 3-bit (Q3_K_M) | usable | 133 GB | 134 GB | 136 GB | 145 GB |
| 4-bit (Q4_K_M) | good, most popular | 167 GB | 167 GB | 169 GB | 178 GB |
| 5-bit (Q5_K_M) | very good | 194 GB | 194 GB | 197 GB | 206 GB |
| 6-bit (Q6_K) | near lossless | 223 GB | 224 GB | 226 GB | 235 GB |
| 8-bit (Q8_0) | effectively lossless | 289 GB | 289 GB | 292 GB | 301 GB |
| 16-bit (F16/BF16) | original weights | 543 GB | 543 GB | 546 GB | 555 GB |
Which GPUs and Macs run DeepSeek V4 Flash?
Best quality quant that fits at 8K context. "Tight" = within 10% of usable memory. Speed is a ceiling from memory bandwidth (real speed is typically 50–80% of it).
Prefer an API?
DeepSeek: DeepSeek V4 Flash 0423 is on OpenRouter at $0.09 per 1M input tokens and $0.18 per 1M output tokens. Price details and history →
Questions
How much VRAM does DeepSeek V4 Flash need?
About 167 GB at 4-bit (Q4_K_M) with an 8K context: 166 GB of weights (estimated), 0.8 GB of KV cache and 0.6 GB of runtime overhead. At 8-bit it needs about 289 GB.
What is the smallest GPU that runs DeepSeek V4 Flash?
No single consumer GPU in our table fits it at Q4_K_M with 8K context; the smallest option that does is the Mac M3 Ultra 256GB (256 GB).
Can I run DeepSeek V4 Flash on a Mac?
Yes. At Q4_K_M with 8K context it fits on a Mac M3 Ultra 256GB or larger, using the default macOS GPU memory limit.
Model facts
| Hugging Face repo | deepseek-ai/DeepSeek-V4-Flash |
| Parameters | 290,944,616,402 (291B) · ~14B active (computed from config) |
| Layers / hidden size | 43 / 4096 |
| Attention | GQA; 64 heads, 1 KV heads, head dim 512 |
| KV cache per token (FP16) | 96.8 KiB |
| Max context | 1,048,576 tokens |
| License | mit |
| Released on HF | Apr 22, 2026 |
| Downloads (30d) / likes | 1,180,002 / 2,403 |
Sources: Hugging Face model API · config.json (deepseek-ai/DeepSeek-V4-Flash) · GPU specs from vendor pages (list). Data fetched Nov 17, 2026. Spot an error? Report it.