← All comparisons
DeepSeek V3.1 Terminus (Non-reasoning) vs Qwen2 Instruct 72B
DeepSeek vs Alibaba — side-by-side benchmark comparison
| DeepSeek V3.1 Terminus (Non-reasoning) | Qwen2 Instruct 72B | |
|---|---|---|
| Intelligence Index | 28.5 | 11.7 |
| Coding Index | 31.9 | — |
| Math Index | 53.7 | — |
| Output speed (tok/s) | 0.0 | 0.0 |
| Blended price ($/1M) | $0.45 | $0.00 |
| Time to first token (s) | 0.00s | 0.00s |
| aime | — | 14.7% |
| aime 25 | 53.7% | — |
| artificial analysis coding index | 31.90 | — |
| artificial analysis intelligence index | 28.50 | 11.70 |
| artificial analysis math index | 53.70 | — |
| gpqa | 75.1% | 37.1% |
| hle | 8.4% | 3.7% |
| ifbench | 41.2% | — |
| lcr | 43.3% | — |
| livecodebench | 52.9% | 15.9% |
| math 500 | — | 70.1% |
| mmlu pro | 83.6% | 62.2% |
| scicode | 32.1% | 22.9% |
| tau2 | 37.1% | — |
| terminalbench hard | 31.8% | — |
Benchmark data from Artificial Analysis.