← All comparisons
Qwen3.5 2B (Non-reasoning) vs Gemini 2.5 Flash (Non-reasoning)
Alibaba vs Google — side-by-side benchmark comparison
| Qwen3.5 2B (Non-reasoning) | Gemini 2.5 Flash (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 14.7 | 20.6 |
| Coding Index | 4.9 | 17.8 |
| Math Index | — | 60.3 |
| Output speed (tok/s) | 272.0 | 214.5 |
| Blended price ($/1M) | $0.04 | $0.85 |
| Time to first token (s) | 0.27s | 0.51s |
| aime | — | 50.0% |
| aime 25 | — | 60.3% |
| artificial analysis coding index | 4.90 | 17.80 |
| artificial analysis intelligence index | 14.70 | 20.60 |
| artificial analysis math index | — | 60.30 |
| gpqa | 43.8% | 68.3% |
| hle | 4.9% | 5.1% |
| ifbench | 29.1% | 39.0% |
| lcr | 13.7% | 45.9% |
| livecodebench | — | 49.5% |
| math 500 | — | 93.2% |
| mmlu pro | — | 80.9% |
| scicode | 7.2% | 29.1% |
| tau2 | 81.6% | 14.9% |
| terminalbench hard | 3.8% | 12.1% |
Benchmark data from Artificial Analysis.