ProLLM - Entity Extraction — leaderboard
ProLLM benchmark evaluating LLMs on structured entity extraction from unstructured text. Measured by F1 score.
Metric: Score (%). Source: www.prollm.ai. Status: saturated. 48 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Grok 3 | 85.5 |
| 2 | GPT-4o | 85.4 |
| 3 | Mistral Small 3 | 85.3 |
| 4 | MiniMax-Text-01 | 85.1 |
| 5 | O1 | 84.8 |
| 6 | Nova Pro | 84.5 |
| 7 | GPT-5 Nano | 84.2 |
| 8 | Gemma 2 9B (IT) | 83.9 |
| 9 | DeepSeek V3 | 83.9 |
| 10 | Gemini 2.0 Flash | 83.8 |
| 11 | GPT-4o Mini | 83.6 |
| 12 | Kimi K2 | 83.3 |
| 13 | O3 Mini (Medium) | 83.1 |
| 14 | Mistral Small 3.1 | 82.8 |
| 15 | GPT-4.1 | 82.7 |
Interactive version: theaggregate.ai/benchmark?slug=prollm-entity-extraction · How the rankings work · Data refreshed daily, snapshot 2026-07-22.