← All comparisons
DeepSeek R1 Distill Qwen 14B vs Qwen3 30B A3B 2507 Instruct
DeepSeek vs Alibaba — side-by-side benchmark comparison
| DeepSeek R1 Distill Qwen 14B | Qwen3 30B A3B 2507 Instruct | |
|---|---|---|
| Intelligence Index | 15.8 | 15.0 |
| Coding Index | — | 14.2 |
| Math Index | 55.7 | 66.3 |
| Output speed (tok/s) | 0.0 | 102.1 |
| Blended price ($/1M) | $0.00 | $0.21 |
| Time to first token (s) | 0.00s | 0.98s |
| aime | 66.7% | 72.7% |
| aime 25 | 55.7% | 66.3% |
| artificial analysis coding index | — | 14.20 |
| artificial analysis intelligence index | 15.80 | 15.00 |
| artificial analysis math index | 55.70 | 66.30 |
| gpqa | 48.4% | 65.9% |
| hle | 4.4% | 6.8% |
| ifbench | 22.1% | 33.1% |
| lcr | 7.0% | 22.7% |
| livecodebench | 37.6% | 51.5% |
| math 500 | 94.9% | 97.5% |
| mmlu pro | 74.0% | 77.7% |
| scicode | 23.9% | 30.4% |
| tau2 | — | 10.2% |
| terminalbench hard | — | 6.1% |
Benchmark data from Artificial Analysis.