← All comparisons
Llama 3.2 Instruct 1B vs Claude 4.1 Opus (Reasoning)
Meta vs Anthropic — side-by-side benchmark comparison
| Llama 3.2 Instruct 1B | Claude 4.1 Opus (Reasoning) | |
|---|---|---|
| Intelligence Index | 6.3 | 42.0 |
| Coding Index | 0.6 | 36.5 |
| Math Index | 0.0 | 80.3 |
| Output speed (tok/s) | 89.7 | 44.5 |
| Blended price ($/1M) | $0.05 | $32.81 |
| Time to first token (s) | 0.59s | 8.55s |
| aime | 0.0% | — |
| aime 25 | 0.0% | 80.3% |
| artificial analysis coding index | 60.0% | 36.50 |
| artificial analysis intelligence index | 6.30 | 42.00 |
| artificial analysis math index | 0.0% | 80.30 |
| gpqa | 19.6% | 80.9% |
| hle | 5.3% | 11.9% |
| ifbench | 22.8% | 55.4% |
| lcr | 5.0% | 66.3% |
| livecodebench | 1.9% | 65.4% |
| math 500 | 14.0% | — |
| mmlu pro | 20.0% | 88.0% |
| scicode | 1.7% | 40.9% |
| tau2 | 0.0% | 71.4% |
| terminalbench hard | 0.0% | 34.3% |
Benchmark data from Artificial Analysis.