Act2Intention - Intention Understanding Semantic Accuracy: leaderboard
Metric: Embedding cosine similarity to reference intention (0-1). Source: arxiv.org. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 1.5 Pro | 0.32 |
| 2 | Claude 3.5 Sonnet | 0.29 |
| 3 | Qwen 2.5 72B | 0.27 |
| 4 | GPT-4o | 0.25 |
| 5 | Qwen 2.5 7B | 0.24 |
| 6 | Claude 3.5 Haiku | 0.24 |
| 7 | GPT-4o Mini | 0.21 |
| 8 | GPT-3.5 Turbo | 0.17 |
| 9 | Llama 3.1 8B | 0.11 |
| 10 | Mistral 7B | 0.11 |
Interactive version: theaggregate.ai/benchmark?slug=act2intention-intention-understanding-semantic-accuracy · How It Works · Data refreshed daily, snapshot 2026-09-19.