← All benchmarks
DeepSeek V4 Flash (Reasoning, Max Effort)
DeepSeek · Released 2026-04-24 · Independent benchmark scores from Artificial Analysis
Intelligence Index
40.3
Output speed
125.4
tokens/sec
Blended price
$0.17
per 1M tokens
Time to first token
0.95s
Model metadata
| Model ID | d8ddb241-b3e4-4c25-a6a3-72eb1b30c541 |
|---|---|
| Slug | deepseek-v4-flash |
| Release date | 2026-04-24 |
| Provider | DeepSeek |
| Provider slug | deepseek |
| Output speed | 125.4 tok/s |
| Time to first token | 0.95s |
| Time to first answer | 45.72s |
| Blended price ($/1M) | $0.17 |
| Input price ($/1M) | $0.14 |
| Output price ($/1M) | $0.28 |
All evaluation scores (17)
| aime | — |
|---|---|
| aime 25 | — |
| artificial analysis coding index | 56.20 |
| artificial analysis intelligence index | 40.30 |
| artificial analysis math index | — |
| gpqa | 89.4% |
| hle | 32.1% |
| ifbench | 79.2% |
| lcr | 63.0% |
| livecodebench | — |
| math 500 | — |
| mmlu pro | — |
| scicode | 44.9% |
| tau banking | 22.9% |
| tau2 | 95.0% |
| terminalbench hard | 35.6% |
| terminalbench v2 1 | 61.8% |
Source: artificialanalysis.ai/models/deepseek-v4-flash
Benchmark data from Artificial Analysis.