CRMArena - Overall (Act): leaderboard

Metric: Mean task score (%). Source: arxiv.org. Saturation forecast: Estimated already saturated. 7 models tracked.

Top models

#ModelScore
1Claude 3.5 Sonnet (20240620)37.4
2GPT-4o (2024-08-06)29.4
3Llama 3.1 405B Instruct22.2
4Llama 3.1 70B Instruct18.6
5GPT-4o Mini16.7
6Claude 3 Sonnet (20240229)16.6
7Llama 3.1 8B Instruct2.5

Interactive version: theaggregate.ai/benchmark?slug=crmarena-overall-act · How It Works · Data refreshed daily, snapshot 2026-10-09.