QCalEval - Outcome Classification: leaderboard

Four-way classification of the experimental outcome (expected, suboptimal parameters, anomalous, apparatus issue) on QCalEval (2026), 243 quantum computing calibration examples from 22 experiment families of superconducting-qubit and neutral-atom experiments, answered zero-shot with the experiment background. Accuracy in percent; higher is better.

Metric: Accuracy (%) of the four-way experimental outcome label (expected behavior, suboptimal parameters, anomalous behavior, apparatus issue), on the 243 quantum-calibration examples 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 March 2027. 18 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Ising-Calibration-1-35B-A3B67.1
2Gemini 3.1 Pro (Preview)57.2#99
3Gemma 4 31B (IT)54.3#350
4Gemini 3.1 Flash Lite53.5#371
5GPT-5.452.7#135
6Claude Opus 4.649#113
7Claude Sonnet 4.648.6#173
8Qwen 3.5 27B45.7#375
9Qwen 3.5 122B A10B44#346
10Qwen 3.5 397B A17B42.8#264
11Qwen 3.5 35B A3B39.9#492
12GPT-5.4 Mini39.5#406
13Qwen 3.5 9B37.9#828
14InternVL3-78B37#759
15Claude Haiku 4.536.6#583

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=qcaleval-outcome-classification · How It Works · Data refreshed daily, snapshot 2026-11-21.