← All comparisons
Llama 3.1 Instruct 405B vs Claude 4.1 Opus (Non-reasoning)
Meta vs Anthropic — side-by-side benchmark comparison
| Llama 3.1 Instruct 405B | Claude 4.1 Opus (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 17.4 | 36.0 |
| Coding Index | 14.5 | — |
| Math Index | 3.0 | — |
| Output speed (tok/s) | 37.5 | 44.7 |
| Blended price ($/1M) | $3.69 | $32.81 |
| Time to first token (s) | 0.63s | 1.63s |
| aime | 21.3% | — |
| aime 25 | 3.0% | — |
| artificial analysis coding index | 14.50 | — |
| artificial analysis intelligence index | 17.40 | 36.00 |
| artificial analysis math index | 3.00 | — |
| gpqa | 51.5% | — |
| hle | 4.2% | — |
| ifbench | 39.0% | — |
| lcr | 24.3% | — |
| livecodebench | 30.5% | — |
| math 500 | 70.3% | — |
| mmlu pro | 73.2% | — |
| scicode | 29.9% | — |
| tau2 | 19.0% | — |
| terminalbench hard | 6.8% | — |
Benchmark data from Artificial Analysis.