← All comparisons
LFM2.5-VL-1.6B vs Claude 4 Sonnet (Non-reasoning)
Liquid AI vs Anthropic — side-by-side benchmark comparison
| LFM2.5-VL-1.6B | Claude 4 Sonnet (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 6.2 | 33.0 |
| Coding Index | 1.0 | 30.6 |
| Math Index | — | 38.0 |
| Output speed (tok/s) | 0.0 | 53.9 |
| Blended price ($/1M) | $0.00 | $6.56 |
| Time to first token (s) | 0.00s | 0.82s |
| aime | — | 40.7% |
| aime 25 | — | 38.0% |
| artificial analysis coding index | 100.0% | 30.60 |
| artificial analysis intelligence index | 6.20 | 33.00 |
| artificial analysis math index | — | 38.00 |
| gpqa | 28.9% | 68.3% |
| hle | 5.1% | 4.0% |
| ifbench | 33.1% | 45.4% |
| lcr | 0.0% | 44.3% |
| livecodebench | — | 44.9% |
| math 500 | — | 93.4% |
| mmlu pro | — | 83.7% |
| scicode | 3.0% | 37.3% |
| tau2 | 8.5% | 52.3% |
| terminalbench hard | 0.0% | 27.3% |
Benchmark data from Artificial Analysis.