QCalEval (In-Context): leaderboard
Metric: Mean score (0-100) over significance analysis, parameter extraction and calibration diagnosis with in-context demonstrations, 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 | Gemini 3.1 Pro (Preview) | 85.2 |
| 2 | Claude Opus 4.6 | 85.1 |
| 3 | Gemma 4 31B (IT) | 81.2 |
| 4 | GPT-5.4 | 78.4 |
| 5 | Gemini 3.1 Flash Lite | 78.1 |
| 6 | Claude Sonnet 4.6 | 75.9 |
| 7 | GPT-5.4 Mini | 66.1 |
| 8 | Claude Haiku 4.5 | 66 |
| 9 | InternVL3-38B | 56.9 |
| 10 | Qwen 3.5 27B | 53 |
| 11 | Qwen 3.5 397B A17B | 48 |
| 12 | InternVL3-78B | 47 |
| 13 | Qwen 3.5 122B A10B | 44.6 |
| 14 | Qwen 3.5 35B A3B | 43.9 |
| 15 | Qwen 3.5 9B | 43.2 |
Interactive version: theaggregate.ai/benchmark?slug=qcaleval-in-context · How It Works · Data refreshed daily, snapshot 2026-10-07.