OpenAI user-flagged hallucination eval
Production conversations users flagged for factual errors. Measures the rate of incorrect claims.
- Made by
- OpenAI
- Kind
- Hallucination rates
- Scored
- Graded by a model
- Answers
- Open-ended
- Tasks
- Kept private
- Measures
- Propensity
- Behavior
- Hallucination
- Language
- English
- Interaction
- Multi-turn
- Items from
- Real work and use
- Refresh
- One fixed set
- Run by
- A model developer
- In views
- Lab internal evals
- Strongest model tested
- GPT-6 Sol #13 on the capability index
- Last in a system card
- November 2026
- Last model added
- November 2026
GPT-5.5 leads with 9.20% among models released by September 2026
Best score per model as labs report them in their own cards, drawn striped. Labs pick their own settings, so these compare less cleanly than independent runs.
The newest model with a score here came out in September 2026, and nobody has reported a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6.1 Sol
- and 15 more
- 1 GPT-5.5 9.20%
- 2 GPT-6 Sol 13.5%
- 3 GPT-6 Luna 18.7%
- 4 GPT-5.5 Instant 46.1%
Source: Labs’ model cards
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-5.5 | 9.20% | Apr 2026 | Lab model card | |
| 2 | GPT-6 Sol | 13.5% | Oct 2026 | Lab model card | |
| 3 | GPT-6 Luna | 18.7% | Oct 2026 | Lab model card | |
| 4 | GPT-5.5 Instant | 46.1% | May 2026 | Lab model card |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Labs’ model cards
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Labs’ model cards
Labs report it only in system cards
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark. Some system cards do, below.
19 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. OpenAI accounts for 11 of the 11 releases that named it since 2024.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| Nov 17, 2026 | GPT-6.2 Sol system card | Performance in Cases Flagged by Users, refreshed 1,000-case version, error rates vs simulated response time, chart without printed numbers (Figure 25) | Hallucination | |
| Nov 17, 2026 | GPT-6.2 Astra system card | Performance in Cases Flagged by Users, refreshed 1,000-case version, error rates vs simulated response time, chart without printed numbers (Figure 25) | Hallucination | |
| Oct 29, 2026 | GPT-6 Luna system card addendum | 24.6 | User Flagged Failures, responses with factual errors (claims with errors 7.2%), printed chart label (Figure 3) | Hallucination |
| Oct 7, 2026 | GPT-6 Sol system card addendum | 13.5 | User Flagged Failures, responses with factual errors (claims with errors 3.5%), printed chart label (Figure 3) | Hallucination |
| Oct 7, 2026 | GPT-6 Luna system card addendum | 18.7 | User Flagged Failures, responses with factual errors (claims with errors 5.1%), printed chart label (Figure 3) | Hallucination |
| Sep 29, 2026 | GPT-6.1 Sol system card addendum | Performance in Cases Flagged by Users, error rate vs latency, chart without printed numbers (Figure 5, 6.1) | Hallucination | |
| Sep 22, 2026 | GPT-6 Sol system card | Performance in Cases Flagged by Users, error rate vs latency, chart without printed numbers (Figure 60, A.5.1) | Hallucination | |
| Sep 22, 2026 | GPT-6 Luna system card | Performance in Cases Flagged by Users, error rate vs latency, chart without printed numbers (Figure 60, A.5.1) | Hallucination | |
| Sep 3, 2026 | GPT-6 Astra system card | Performance in Cases Flagged by Users, error rate vs latency, chart without printed numbers (Figure 6, 7.1) | Hallucination | |
| Aug 6, 2026 | GPT-5.6 Sol system card addendum | User Flagged Failures, relative hallucination rate vs GPT-5.3 Instant, % responses with factual error, 0.31x, printed chart label (Figure 2, 6) | Hallucination | |
| Aug 6, 2026 | GPT-5.6 Luna system card addendum | User Flagged Failures, relative hallucination rate vs GPT-5.3 Instant, % responses with factual error, 0.49x, printed chart label (Figure 2, 6) | Hallucination | |
| Jul 9, 2026 | GPT-5.6 Terra system card | Performance in Cases Flagged by Users, error rate vs simulated latency, chart without printed numbers (Figure 4, 6.1) | Hallucination | |
| Jul 9, 2026 | GPT-5.6 Sol system card | Performance in Cases Flagged by Users, error rate vs simulated latency, chart without printed numbers (Figure 4, 6.1) | Hallucination | |
| Jul 9, 2026 | GPT-5.6 Luna system card | Performance in Cases Flagged by Users, error rate vs simulated latency, chart without printed numbers (Figure 4, 6.1) | Hallucination | |
| Jun 25, 2026 | GPT-5.6 Terra system card | Performance in Cases Flagged by Users, error rate vs simulated latency, chart without printed numbers (Figure 4, 6.1) | Hallucination | |
| Jun 25, 2026 | GPT-5.6 Sol system card | Performance in Cases Flagged by Users, error rate vs simulated latency, chart without printed numbers (Figure 4, 6.1) | Hallucination | |
| Jun 25, 2026 | GPT-5.6 Luna system card | Performance in Cases Flagged by Users, error rate vs simulated latency, chart without printed numbers (Figure 4, 6.1) | Hallucination | |
| May 4, 2026 | GPT-5.5 Instant system card | 46.1 | User flagged failures, responses with factual error (claims with factual error 15.8%), printed chart label (Figure 2) | Hallucination |
| Apr 23, 2026 | GPT-5.5 system card | 9.2 | Performance in Cases Flagged by Users, response-level hallucination rate, printed chart label (claim-level 2.0%) (Figure 3, 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%
- 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
- GaRAGe In use Models write long-form RAG answers to 2,366 questions with human-annotated grounding passages. Scoring checks whether answers use the relevant passages.
Amazon Nova Pro 60.7%
- 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%
Data audit
3 of 4 checks pass, 1 warning. Look first at the Confidence check. Likely. Some facts rest on secondary sources. Every benchmark
Catalog record 1 warning 1 pass
- Pass Task access The tasks are private.
- Warning Confidence Likely. Some facts rest on secondary sources.
Scores 1 pass 4 notes
- Note Independent runs Only labs’ own numbers, as expected for an in-house eval. Scores from system cards (5 scores).
- Note Models scored Only 4 models we track.
- Pass Newest score The newest is from November 2026, for GPT-6.2 Astra.
- Note Frontier models None of the top 10 has a score, which is fine for an in-house eval. The strongest tested is GPT-6 Sol (#13).
- Note Headroom Lower is better, so there’s no ceiling to reach.
Leaderboard import 1 note
- Note Importer No public leaderboard to read.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.