QCalEval - Fit Assessment: leaderboard
Metric: Accuracy (%) of the three-way fit reliability label (reliable, unreliable, no fit), on the 243 quantum-calibration plots of QCalEval (22 experiment families, superconducting qubits and neutral atoms), zero-shot with the family background text, greedy decoding; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 18 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 84.4 |
| 2 | Gemma 4 31B (IT) | 82.7 |
| 3 | Gemini 3.1 Flash Lite | 82.7 |
| 4 | Claude Sonnet 4.6 | 76.5 |
| 5 | Claude Opus 4.6 | 76.1 |
| 6 | Qwen 3.5 27B | 56.4 |
| 7 | GPT-5.4 | 54.7 |
| 8 | Qwen 3.5 35B A3B | 52.7 |
| 9 | Qwen 3.5 397B A17B | 50.6 |
| 10 | Qwen 3.5 122B A10B | 50.2 |
| 11 | Qwen 3.5 9B | 49.8 |
| 12 | Claude Haiku 4.5 | 48.6 |
| 13 | InternVL3-78B | 42.8 |
| 14 | GPT-5.4 Mini | 42 |
| 15 | InternVL3-38B | 33.7 |
Interactive version: theaggregate.ai/benchmark?slug=qcaleval-fit-assessment · How It Works · Data refreshed daily, snapshot 2026-10-07.