DeepSeek V4 Flash (FP4)
DeepSeek V4 Flash — 284B-parameter Mixture-of-Experts (~13B active per token) with 1M-token context, Compressed Sparse Attention, and configurable Non-think / Think-High / Think-Max reasoning modes. On the ds4 engine this is the July 2026 0731 re-post-trained checkpoint, which is substantially stronger at agentic and coding work than the original release. The llama.cpp (native FP4+FP8) and MLX sources still carry the original V4 Flash weights — upstream has not republished those builds for 0731 — so pick the ds4 source to get the newer checkpoint. Frontier-tier on-device reasoning for workstations and Mac Studio-class hardware — needs ~200 GB of usable memory.
| At a glance | |
|---|---|
| Made by | DeepSeek |
| Download size | 164.6 GB |
| Parameters | 284B |
| Quantization | FP4-FP8 |
| Context window | 1000K tokens |
| Uses tools | Yes |
| Best for | Reasoning |
| Runs on | llama.cpp, MLX, DwarfStar |
| License | MIT |
| Added to the catalog | 2026-08-01 |
Gezel downloads this model the first time a gezel needs it, checks the file against its published hash, and applies tuned settings for it.