# DeepSeek V4 Flash: hardware requirements
> DeepSeek V4 Flash (DeepSeek) is a 284B (13B active) mixture-of-experts model with a 1.0M-token context window, released under the MIT license. It needs about 168GB of memory at FP4 with 8K context.
DeepSeek V4 Flash is the smaller model of DeepSeek's April 2026 V4 generation: 284B total parameters, 13B active, a 1M-token context, and a toggleable thinking mode, refreshed with improved 0731 weights in July. Its compressed attention makes long contexts remarkably cheap, and its expert weights ship natively in FP4, so the full-quality build is only about 162GB. Running it at full quality takes a 256GB Mac Studio or a multi-GPU server with about 170GB of memory; 3-bit dynamic builds of roughly 103GB to 110GB fit 128GB machines with some quality loss. DeepSeek's September V4.1 Flash is larger (552B, including 196B of Engram memory), so V4 Flash remains the easier DeepSeek to run locally.

## Recommended hardware
- **Multi-GPU:** 8× RTX 3090 (192GB), FP4, ~25 tokens/sec
- **Multi-GPU:** 8× RTX 4090 (192GB), FP4, ~27 tokens/sec
- **Multi-GPU:** 8× RTX 5090 (256GB), FP4, ~33 tokens/sec
- **Mac / unified memory:** M3 Ultra 512GB (512GB), FP4, ~32 tokens/sec

## Memory by quantization
| Quant | Weights | Total @ 8K ctx | Total @ 32K ctx |
| --- | --- | --- | --- |
| FP4 | 162.0GB | 167.5GB | 167.6GB |

## GPUs that run it fully in memory (fastest first)
| GPU | Memory | Quant | Speed | Price |
| --- | --- | --- | --- | --- |
| M5 Ultra 512GB | 512GB | FP4 | ~45 tok/s | n/a |
| M5 Ultra 256GB | 256GB | FP4 | ~45 tok/s | ~$9,499 |
| M3 Ultra 512GB | 512GB | FP4 | ~32 tok/s | $9,499 MSRP |

## How to run it
- llama.cpp: `llama-server -hf unsloth/DeepSeek-V4-Flash-GGUF -c 8192 -ngl 99`
- Weights: https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash

## Strengths
- Near-frontier reasoning and coding with only 13B active parameters
- 1M-token context with a very small KV cache
- Native FP4 experts: the full-quality build is about 162GB
- MIT license

## Weaknesses
- Needs about 170GB for full quality; 128GB machines must drop to 3-bit
- Text only (vision is in a separate experimental release)
- Runtime support for its compressed attention arrived months after launch and is still maturing
Full page: https://llamabuilds.ai/models/deepseek-v4-flash · Hardware guide: https://llamabuilds.ai/models/deepseek-v4-flash/requirements

---
Source: https://llamabuilds.ai/models/deepseek-v4-flash · Updated October 14, 2026 · Llama Builds. Speeds marked ~ are estimates from our bandwidth model (https://llamabuilds.ai/methodology); measured results are labeled as such.