BENCHMARKS
MEASUREMENTS, NOT RECOMMENDATIONS
Benchmarks
Scores retain their source and confidence. They are not merged into an editorial master score.
| Benchmark | Model | Score | Metric | Confidence |
|---|---|---|---|---|
| LiveBench | GLM-5.3 | 71.2 | reported score | LOW |
| LiveBench | DeepSeek V3 0324 | 65.1 | reported score | LOW |
| LiveBench | GPT-OSS 120B | 63.4 | reported score | LOW |