← All comparisons
Trinity Large Thinking vs Qwen3 VL 4B (Reasoning)
Arcee AI vs Alibaba — side-by-side benchmark comparison
| Trinity Large Thinking | Qwen3 VL 4B (Reasoning) | |
|---|---|---|
| Intelligence Index | 31.9 | 13.7 |
| Coding Index | 27.2 | 6.7 |
| Math Index | — | 25.7 |
| Output speed (tok/s) | 171.4 | 0.0 |
| Blended price ($/1M) | $0.40 | $0.00 |
| Time to first token (s) | 0.67s | 0.00s |
| aime | — | — |
| aime 25 | — | 25.7% |
| artificial analysis coding index | 27.20 | 6.70 |
| artificial analysis intelligence index | 31.90 | 13.70 |
| artificial analysis math index | — | 25.70 |
| gpqa | 75.2% | 49.4% |
| hle | 14.7% | 4.4% |
| ifbench | 56.3% | 36.6% |
| lcr | 33.0% | 21.3% |
| livecodebench | — | 32.0% |
| math 500 | — | — |
| mmlu pro | — | 70.0% |
| scicode | 36.1% | 17.1% |
| tau2 | 90.1% | 15.5% |
| terminalbench hard | 22.7% | 1.5% |
Benchmark data from Artificial Analysis.