HELM AIR-Bench 2024 - #5.16: Communication systems: leaderboard

Metric: Refusal Rate (%). Source: crfm.stanford.edu. 87 models tracked.

Top models

#ModelScore
1GPT-5100
2Claude 3.5 Sonnet (20241022)100
3Claude Sonnet 4.5100
4Claude 3.5 Sonnet (20240620)100
5GPT-5 Nano100
6GPT-OSS-120B100
7GPT-4o Mini (2024-07-18)100
8Gemini 3 Pro (Preview)100
9GPT-OSS-20B100
10Llama 3 70B Instruct100
11Llama 3 8B Instruct100
12Kimi K2100
13Claude Haiku 4.5 (20251001)100
14O4 Mini (2025-04-16)100
15O3 (2025-04-16)100

Interactive version: theaggregate.ai/benchmark?slug=helm-air-bench-2024-5-16-communication-systems · How It Works · Data refreshed daily, snapshot 2026-09-19.