xAI Factuality (Hallucination) eval
Internal measure of unsupported or fabricated claims in single responses to information-seeking queries.
- Made by
- xAI (SpaceXAI)
- Kind
- Hallucination rates
- Scored
- Graded by a model
- Answers
- Open-ended
- Tasks
- Kept private
- Strongest model tested
- Grok 4.6 #28 on the capability index
- Last in a system card
- August 2026
- Last model added
- August 2026
- Scores refreshed
- November 12th, 2026
Grok 4.5 leads with 0.98%
Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.
Grok 4.1 leads with 4.22% across all independent runs
Best score per model across the official leaderboard, Epoch AI’s own runs and any board Epoch AI copied. Each bar and row names its source.
Grok 4.5 leads with 0.98%, claims included
Best score per model from every source, including labs’ own numbers and claims Epoch AI collected from reports and papers. Labs pick their own settings, so a claim can beat an independent run of the same model.
Source: Official xAI Factuality (Hallucination) eval leaderboard
- 1 Grok 4.5 0.98%
- 2 Grok 4.6 1.70%
- 3 Grok 4.1 Non-reasoning, web search 4.22%
- 4 Grok 4 Fast Non-reasoning, web search 12.1%
Sources: Labs’ model cards and official xAI Factuality (Hallucination) eval leaderboard
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Grok 4.5 | 0.98% | Jul 2026 | Lab model card | |
| 2 | Grok 4.6 | 1.70% | Aug 2026 | Lab model card |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Grok 4.1 | 4.22% | Non-reasoning, web search | Nov 2025 | Official leaderboard |
| 2 | Grok 4 Fast | 12.1% | Non-reasoning, web search | Sep 2025 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Grok 4.5 | 0.98% | Jul 2026 | Lab model card | |
| 2 | Grok 4.6 | 1.70% | Aug 2026 | Lab model card | |
| 3 | Grok 4.1 | 4.22% | Non-reasoning, web search | Nov 2025 | Official leaderboard |
| 4 | Grok 4 Fast | 12.1% | Non-reasoning, web search | Sep 2025 | Official leaderboard |
The benchmark's own leaderboard has 2 models. Switch the scores above to rank them together.
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Sources: Official xAI Factuality (Hallucination) eval leaderboard and labs’ model cards
Labs report it only in system cards
We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark. Some system cards do, below.
2 system cards report it, from 1 lab
Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| Aug 12, 2026 | Grok 4.6 model card | 1.7 | Factuality (Hallucination), hallucination rate (lower is better), high (6.1) | Hallucination |
| Jul 14, 2026 | Grok 4.5 model card | 0.98 | Factuality, single-turn hallucination rate (lower is better), high (6.1) | Hallucination |
More in Factuality
- SimpleVQA In use Models answer 2,025 short factual questions about images across 9 tasks and 9 topics. An LLM judge grades answers against fixed references.
Qwen3.7-Plus 81.7% - HallusionBench In use Models answer 1,129 yes or no questions about 346 images, built in pairs to expose language hallucination and visual illusion. Scored by accuracy.
Keye-VL-2.0-30B-A3B 73.5% - SimpleQA Verified In use 1,000 human-verified, de-duplicated short-answer factual questions measuring parametric knowledge.
Gemini 3.1 Pro 77.5% - Kimi Hallu Bench In use Models answer hard prompts drawn from practical work tasks. Fact-checking agents score the share of claims that are hallucinated. GPT-6 Astra 4.20%
- Internal adversarially-designed factuality evaluations In use Models answer adversarial questions in nine categories such as history, math and code. Scored on agreement with human ideal responses. GPT-6 Luna 7.60%
- AA-Omniscience v1.1 In use Models answer 3,000 knowledge questions in six domains without tools and may abstain. The index adds correct answers and subtracts wrong ones.
Inkling 2 Small 16.1