Skip to content

The Aggregate

Unified LLM rankings, daily

Models What's New Trends Benchmarks Skill Maps
How It Works Guesswork Metrics
The Aggregate
Loading data...

Vector Eval - AgentHarm — leaderboard

Metric: Average Harm Score (%) (lower is better). Source: huggingface.co. 5 models tracked.

Top models

#ModelScore
1O18.78
2O3 Mini12.42
3Claude 3.5 Sonnet14.77
4Gemini 1.5 Pro28.99
5GPT-4o49.95

Interactive version: theaggregate.ai/benchmark?slug=vector-eval-agentharm · How the rankings work · Data refreshed daily, snapshot 2026-07-22.

Built by Mikhail Doroshenko — AI researcher, co-author of Humanity’s Last Exam. Independent project; no affiliation with any AI lab.