← All comparisons
Llama 3 Instruct 70B vs Claude 4.1 Opus (Non-reasoning)
Meta vs Anthropic — side-by-side benchmark comparison
| Llama 3 Instruct 70B | Claude 4.1 Opus (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 8.9 | 36.0 |
| Coding Index | 6.8 | — |
| Math Index | — | — |
| Output speed (tok/s) | 46.2 | 44.7 |
| Blended price ($/1M) | $1.18 | $32.81 |
| Time to first token (s) | 0.68s | 1.63s |
| aime | 0.0% | — |
| aime 25 | — | — |
| artificial analysis coding index | 6.80 | — |
| artificial analysis intelligence index | 8.90 | 36.00 |
| artificial analysis math index | — | — |
| gpqa | 37.9% | — |
| hle | 4.4% | — |
| ifbench | 37.1% | — |
| lcr | 0.0% | — |
| livecodebench | 19.8% | — |
| math 500 | 48.3% | — |
| mmlu pro | 57.4% | — |
| scicode | 18.9% | — |
| tau2 | 0.0% | — |
| terminalbench hard | 0.8% | — |
Benchmark data from Artificial Analysis.