DeepSeek-V4-Flash
DeepSeek · 2 models
- Sizes
- 291B–304B
- Downloads, last 30 days
- 3.2M
- Top Arena score
- 1432
- Built on them
- 428
Models
| Model | Params | Score |
|---|---|---|
| 291B | 1432 | |
| 304B | — |
API tokens per day
Last 7 days5.2T tokens−4% compared with the 7 days beforeSource: OpenRouter (openrouter.ai/rankings), as of 2026-11-17.
Built on these models
- Quantizations
- 342
- Fine-tunes
- 71
- Adapters
- 14
- Merges
- 1
Discussed, past year
- DeepSeek V4-Flash локально | 288GB VRAM на четырёх GPU

DeepSeek-V4-Flash-0731· Aug 204.4K likes
- DeepSeek V4 Flash 0731

DeepSeek-V4-Flash-0731· Aug 7796 points
- DeepSeek-V4-Flash Update

DeepSeek-V4-Flash-0731· Jul 31745 points
- DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis

DeepSeek-V4-Flash-0731· Jul 31594 points
- DeepSeek V4 Flash on a Single AMD MI300X

DeepSeek-V4-Flash-0731· Aug 4382 points
- DeepSeek-V4-Flash means LLM steering is interesting again

DeepSeek-V4-Flash-0731· May 16280 points