← All comparisons
Exaone 4.0 1.2B (Reasoning) vs Qwen3.5 397B A17B (Non-reasoning)
LG AI Research vs Alibaba — side-by-side benchmark comparison
| Exaone 4.0 1.2B (Reasoning) | Qwen3.5 397B A17B (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 8.3 | 40.1 |
| Coding Index | 3.1 | 37.4 |
| Math Index | 50.3 | — |
| Output speed (tok/s) | 0.0 | 53.5 |
| Blended price ($/1M) | $0.00 | $1.35 |
| Time to first token (s) | 0.00s | 1.85s |
| aime | — | — |
| aime 25 | 50.3% | — |
| artificial analysis coding index | 3.10 | 37.40 |
| artificial analysis intelligence index | 8.30 | 40.10 |
| artificial analysis math index | 50.30 | — |
| gpqa | 51.5% | 86.1% |
| hle | 5.8% | 18.8% |
| ifbench | 23.0% | 51.6% |
| lcr | 0.0% | 58.0% |
| livecodebench | 51.6% | — |
| math 500 | — | — |
| mmlu pro | 58.8% | — |
| scicode | 9.3% | 41.1% |
| tau2 | 16.4% | 83.9% |
| terminalbench hard | 0.0% | 35.6% |
Benchmark data from Artificial Analysis.