← All comparisons
GPT-4 vs Claude 3.5 Sonnet (June '24)
OpenAI vs Anthropic — side-by-side benchmark comparison
| GPT-4 | Claude 3.5 Sonnet (June '24) | |
|---|---|---|
| Intelligence Index | 12.8 | 14.2 |
| Coding Index | 13.1 | 26.0 |
| Math Index | — | — |
| Output speed (tok/s) | 35.8 | 0.0 |
| Blended price ($/1M) | $37.50 | $6.56 |
| Time to first token (s) | 0.79s | 0.00s |
| aime | — | 9.7% |
| aime 25 | — | — |
| artificial analysis coding index | 13.10 | 26.00 |
| artificial analysis intelligence index | 12.80 | 14.20 |
| artificial analysis math index | — | — |
| gpqa | — | 56.0% |
| hle | — | 3.7% |
| ifbench | — | — |
| lcr | — | — |
| livecodebench | — | — |
| math 500 | — | 69.5% |
| mmlu pro | — | 75.1% |
| scicode | — | 31.6% |
| tau2 | — | — |
| terminalbench hard | — | — |
Benchmark data from Artificial Analysis.