← All comparisons
Llama 3.1 Instruct 8B vs Qwen3.6 Max Preview
Meta vs Alibaba — side-by-side benchmark comparison
| Llama 3.1 Instruct 8B | Qwen3.6 Max Preview | |
|---|---|---|
| Intelligence Index | 11.8 | 51.8 |
| Coding Index | 4.9 | 44.9 |
| Math Index | 4.3 | — |
| Output speed (tok/s) | 212.7 | 38.7 |
| Blended price ($/1M) | $0.10 | $2.92 |
| Time to first token (s) | 0.45s | 1.96s |
| aime | 7.7% | — |
| aime 25 | 4.3% | — |
| artificial analysis coding index | 4.90 | 44.90 |
| artificial analysis intelligence index | 11.80 | 51.80 |
| artificial analysis math index | 4.30 | — |
| gpqa | 25.9% | 88.8% |
| hle | 5.1% | 28.9% |
| ifbench | 28.6% | 76.6% |
| lcr | 15.7% | 69.7% |
| livecodebench | 11.6% | — |
| math 500 | 51.9% | — |
| mmlu pro | 47.6% | — |
| scicode | 13.2% | 46.9% |
| tau2 | 16.4% | 95.9% |
| terminalbench hard | 0.8% | 43.9% |
Benchmark data from Artificial Analysis.