← All benchmarks
DeepSeek V4 Flash 0731 (Reasoning, Max Effort)
DeepSeek · Released 2026-07-31 · Independent benchmark scores from Artificial Analysis
Intelligence Index
34.5
Output speed
239.4
tokens/sec
Blended price
$0.66
per 1M tokens
Time to first token
0.78s
Model metadata
| Model ID | fe4c0848-e284-4e52-a79d-cdc28392f1a9 |
|---|---|
| Slug | deepseek-v4-flash |
| Release date | 2026-07-31 |
| Provider | DeepSeek |
| Provider slug | deepseek |
| Output speed | 239.4 tok/s |
| Time to first token | 0.78s |
| Time to first answer | 9.14s |
| Blended price ($/1M) | $0.66 |
| Input price ($/1M) | $0.44 |
| Output price ($/1M) | $1.32 |
All evaluation scores (17)
| aime | — |
|---|---|
| aime 25 | — |
| artificial analysis coding index | 69.10 |
| artificial analysis intelligence index | 34.50 |
| artificial analysis math index | — |
| gpqa | 90.8% |
| hle | 38.6% |
| ifbench | — |
| lcr | 79.7% |
| livecodebench | — |
| math 500 | — |
| mmlu pro | — |
| scicode | 50.3% |
| tau banking | 39.4% |
| tau2 | — |
| terminalbench hard | — |
| terminalbench v2 1 | 78.7% |
Source: artificialanalysis.ai/models/deepseek-v4-flash
Benchmark data from Artificial Analysis.