← All comparisons
Llama 4 Maverick vs Qwen3 Omni 30B A3B Instruct
Meta vs Alibaba — side-by-side benchmark comparison
| Llama 4 Maverick | Qwen3 Omni 30B A3B Instruct | |
|---|---|---|
| Intelligence Index | 18.4 | 10.7 |
| Coding Index | 15.6 | 7.2 |
| Math Index | 19.3 | 52.3 |
| Output speed (tok/s) | 119.7 | 108.9 |
| Blended price ($/1M) | $0.47 | $0.43 |
| Time to first token (s) | 0.64s | 1.04s |
| aime | 39.0% | — |
| aime 25 | 19.3% | 52.3% |
| artificial analysis coding index | 15.60 | 7.20 |
| artificial analysis intelligence index | 18.40 | 10.70 |
| artificial analysis math index | 19.30 | 52.30 |
| gpqa | 67.1% | 62.0% |
| hle | 4.8% | 5.1% |
| ifbench | 43.0% | 31.2% |
| lcr | 46.0% | 0.0% |
| livecodebench | 39.7% | 42.2% |
| math 500 | 88.9% | — |
| mmlu pro | 80.9% | 72.5% |
| scicode | 33.1% | 18.6% |
| tau2 | 17.8% | 16.4% |
| terminalbench hard | 6.8% | 1.5% |
Benchmark data from Artificial Analysis.