RHELM (Long Context) - Hallucination: leaderboard

Metric: Accuracy (%; 197 hallucination questions, LLM judge, dialogue history plus external sources, full history in a 1M-token context). Source: arxiv.org. Saturation forecast: Around June 2028. 3 models tracked.

Top models

#ModelScore
1Gemini 2.5 Flash Lite7.6
2GPT-4.1 Mini2
3Qwen2.5-14B-Instruct-1M1

Interactive version: theaggregate.ai/benchmark?slug=rhelm-long-context-hallucination · How It Works · Data refreshed daily, snapshot 2026-09-25.