← All comparisons
Claude 3.5 Haiku vs DeepSeek R1 0528 Qwen3 8B
Anthropic vs DeepSeek — side-by-side benchmark comparison
| Claude 3.5 Haiku | DeepSeek R1 0528 Qwen3 8B | |
|---|---|---|
| Intelligence Index | 18.7 | 16.4 |
| Coding Index | 10.7 | 7.8 |
| Math Index | — | 63.7 |
| Output speed (tok/s) | 0.0 | 0.0 |
| Blended price ($/1M) | $1.75 | $0.00 |
| Time to first token (s) | 0.00s | 0.00s |
| aime | 3.3% | 65.0% |
| aime 25 | — | 63.7% |
| artificial analysis coding index | 10.70 | 7.80 |
| artificial analysis intelligence index | 18.70 | 16.40 |
| artificial analysis math index | — | 63.70 |
| gpqa | 40.8% | 61.2% |
| hle | 3.5% | 5.6% |
| ifbench | 42.8% | 19.9% |
| lcr | 23.3% | 13.0% |
| livecodebench | 31.4% | 51.3% |
| math 500 | 72.1% | 93.2% |
| mmlu pro | 63.4% | 73.9% |
| scicode | 27.4% | 20.4% |
| tau2 | 24.6% | 0.0% |
| terminalbench hard | 2.3% | 1.5% |
Benchmark data from Artificial Analysis.