← All comparisons
Apertus 8B Instruct vs Qwen3.5 397B A17B (Reasoning)
Swiss AI Initiative vs Alibaba — side-by-side benchmark comparison
| Apertus 8B Instruct | Qwen3.5 397B A17B (Reasoning) | |
|---|---|---|
| Intelligence Index | 5.9 | 45.0 |
| Coding Index | 1.4 | 41.3 |
| Math Index | — | — |
| Output speed (tok/s) | 0.0 | 52.1 |
| Blended price ($/1M) | $0.13 | $1.35 |
| Time to first token (s) | 0.00s | 1.81s |
| aime | — | — |
| aime 25 | — | — |
| artificial analysis coding index | 1.40 | 41.30 |
| artificial analysis intelligence index | 5.90 | 45.00 |
| artificial analysis math index | — | — |
| gpqa | 25.6% | 89.3% |
| hle | 5.0% | 27.3% |
| ifbench | 22.4% | 78.8% |
| lcr | 0.0% | 65.7% |
| livecodebench | — | — |
| math 500 | — | — |
| mmlu pro | — | — |
| scicode | 4.1% | 42.0% |
| tau2 | 11.4% | 95.6% |
| terminalbench hard | 0.0% | 40.9% |
Benchmark data from Artificial Analysis.