DeepSeek V4.1 Flash (deepseek-flash in the API) is DeepSeek's fast, low-cost model with image input. The weights are open under the MIT license. Prices shown are peak rates; off-peak rates are half.
DeepSeek V4.1 Flash
DeepSeek's fast, low-cost model with vision and a 1M-token context
Checked 2026-11-12 against DeepSeek's documentation
| Provider | DeepSeek |
| Status | Current |
| Context window | 1M tokens |
| Pricing | Open |
| Input price | $0.30/M |
| Output price | $1.20/M |
| Capabilities | textvisioncodefunction-callingreasoning |
| API docs | Docs ↗ |
Benchmarks
Benchmark data not yet available for this model.
Compare with similar models
| DeepSeek V4.1 Flash | DeepSeek R1 | Llama 3.1 405B | DeepSeek V3 | |
|---|---|---|---|---|
| Context | 1M | 66K | 131K | 66K |
| MMLU | — | 90.8 | 88.6 | 88.5 |
| HumanEval | — | — | — | 91.6 |
| MATH | — | 97.3 | 73.5 | 90.2 |
| GPQA | — | 71.5 | — | — |
| Pricing | Open | Open | Open | Open |
| Input $/M | $0.30 | $0.55 | — | $0.27 |