PluginEval - Recall: leaderboard
End-to-end recall on the 2,500 positive queries of PluginEval (2026), a Chinese function-calling benchmark with execution-verified gold calls: the share of queries for which the model calls every required tool with correct arguments, judged against gold annotations. Percent; higher is better.
Metric: Recall (%; share of the 2,500 positive queries where every required tool is called with correct arguments; original difficulty distribution; 3,000 human-verified Chinese function-calling queries over 54 plugins and 63 tools with real API execution (2,500 positive queries requiring grounded calls, 500 adversarial negatives requiring abstention; easy/medium/hard 0.4/0.4/0.2); GPT-5.2 judge anchored to gold annotations). Source: arxiv.org. Saturation forecast: Around February 2027. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.6 | 67.6 | #102 |
| 2 | Gemini 3.1 Pro (Preview) | 64.96 | #94 |
| 3 | DeepSeek V4 Pro | 62.94 | #188 |
| 4 | Qwen 3 235B A22B | 54.82 | #623 |
| 5 | GPT-5.4 | 49.2 | #123 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=plugineval-recall · How It Works · Data refreshed daily, snapshot 2026-11-13.