RHELM (Long Context) - Hallucination: leaderboard
RHELM category (Zhang et al., Microsoft and Renmin University of China, 2026) that poses questions built on misattributed or fabricated details, or on requests that contradict the persona's stated preferences or current state, which the assistant must detect and correct. The model reads the whole history, dialogue and external documents in chronological order, in a 1M-token context; accuracy in percent on 197 questions judged 0 or 1 by an LLM, higher is better.
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 July 2028. 3 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Flash Lite | 7.6 | #958 |
| 2 | GPT-4.1 Mini | 2 | #775 |
| 3 | Qwen2.5-14B-Instruct-1M | 1 | #1417 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=rhelm-long-context-hallucination · How It Works · Data refreshed daily, snapshot 2026-11-12.