← All comparisons
Claude 4.5 Haiku (Non-reasoning) vs Nova 2.0 Lite (medium)
Anthropic vs Amazon — side-by-side benchmark comparison
| Claude 4.5 Haiku (Non-reasoning) | Nova 2.0 Lite (medium) | |
|---|---|---|
| Intelligence Index | 31.0 | 29.7 |
| Coding Index | 29.6 | 23.9 |
| Math Index | 39.0 | 88.7 |
| Output speed (tok/s) | 105.8 | 168.5 |
| Blended price ($/1M) | $2.19 | $0.85 |
| Time to first token (s) | 0.65s | 13.49s |
| aime | — | — |
| aime 25 | 39.0% | 88.7% |
| artificial analysis coding index | 29.60 | 23.90 |
| artificial analysis intelligence index | 31.00 | 29.70 |
| artificial analysis math index | 39.00 | 88.70 |
| gpqa | 64.6% | 76.8% |
| hle | 4.3% | 8.6% |
| ifbench | 42.0% | 68.5% |
| lcr | 43.7% | 58.3% |
| livecodebench | 51.1% | 66.3% |
| math 500 | — | — |
| mmlu pro | 80.0% | 81.3% |
| scicode | 34.4% | 36.8% |
| tau2 | 32.5% | 75.7% |
| terminalbench hard | 27.3% | 17.4% |
Benchmark data from Artificial Analysis.