← All comparisons
DeepSeek V4 Flash (Reasoning, Max Effort) vs Gemini 2.5 Flash Preview (Sep '25) (Reasoning)
DeepSeek vs Google — side-by-side benchmark comparison
| DeepSeek V4 Flash (Reasoning, Max Effort) | Gemini 2.5 Flash Preview (Sep '25) (Reasoning) | |
|---|---|---|
| Intelligence Index | 46.5 | 31.1 |
| Coding Index | 38.7 | 24.6 |
| Math Index | — | 78.3 |
| Output speed (tok/s) | 119.3 | 0.0 |
| Blended price ($/1M) | $0.17 | $0.00 |
| Time to first token (s) | 0.86s | 0.00s |
| aime | — | — |
| aime 25 | — | 78.3% |
| artificial analysis coding index | 38.70 | 24.60 |
| artificial analysis intelligence index | 46.50 | 31.10 |
| artificial analysis math index | — | 78.30 |
| gpqa | 89.4% | 79.3% |
| hle | 32.1% | 12.7% |
| ifbench | 79.2% | 52.3% |
| lcr | 63.0% | 64.3% |
| livecodebench | — | 71.3% |
| math 500 | — | — |
| mmlu pro | — | 84.2% |
| scicode | 44.9% | 40.5% |
| tau2 | 95.0% | 45.6% |
| terminalbench hard | 35.6% | 16.7% |
Benchmark data from Artificial Analysis.