QCalEval (In-Context) - Significance Analysis: leaderboard
Metric: Score (0-100) for the scientific significance analysis, three expert key points per item scored 0, 0.5 or 1 by GPT-5.4 and Gemini 3.1 Pro Preview and averaged, with one analysis demonstration per scenario type, on the QCalEval quantum-calibration plots with in-context demonstrations from the same experiment family (scenario types with a single sample left out), greedy decoding; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 17 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.6 | 84.7 |
| 2 | Gemini 3.1 Pro (Preview) | 81.3 |
| 3 | GPT-5.4 | 81 |
| 4 | Gemma 4 31B (IT) | 80.6 |
| 5 | Gemini 3.1 Flash Lite | 78.5 |
| 6 | Claude Sonnet 4.6 | 77.8 |
| 7 | Claude Haiku 4.5 | 66.1 |
| 8 | GPT-5.4 Mini | 58.8 |
| 9 | InternVL3-38B | 56.2 |
| 10 | InternVL3-78B | 50.5 |
| 11 | Qwen 3.5 27B | 41.8 |
| 12 | Qwen 3.5 397B A17B | 37.4 |
| 13 | Qwen 3.5 122B A10B | 36.1 |
| 14 | Qwen 3.5 35B A3B | 33.4 |
| 15 | Qwen 3.5 9B | 32.8 |
Interactive version: theaggregate.ai/benchmark?slug=qcaleval-in-context-significance-analysis · How It Works · Data refreshed daily, snapshot 2026-10-07.