← All comparisons
Exaone 4.0 1.2B (Non-reasoning) vs Qwen3 VL 30B A3B Instruct
LG AI Research vs Alibaba — side-by-side benchmark comparison
| Exaone 4.0 1.2B (Non-reasoning) | Qwen3 VL 30B A3B Instruct | |
|---|---|---|
| Intelligence Index | 8.1 | 16.0 |
| Coding Index | 2.5 | 14.3 |
| Math Index | 24.0 | 72.3 |
| Output speed (tok/s) | 0.0 | 123.5 |
| Blended price ($/1M) | $0.00 | $0.30 |
| Time to first token (s) | 0.00s | 1.07s |
| aime | — | — |
| aime 25 | 24.0% | 72.3% |
| artificial analysis coding index | 2.50 | 14.30 |
| artificial analysis intelligence index | 8.10 | 16.00 |
| artificial analysis math index | 24.00 | 72.30 |
| gpqa | 42.4% | 69.5% |
| hle | 5.8% | 6.4% |
| ifbench | 25.3% | 33.1% |
| lcr | 0.0% | 23.7% |
| livecodebench | 29.3% | 47.6% |
| math 500 | — | — |
| mmlu pro | 50.0% | 76.4% |
| scicode | 7.4% | 30.8% |
| tau2 | 20.5% | 19.0% |
| terminalbench hard | 0.0% | 6.1% |
Benchmark data from Artificial Analysis.