reasoning benchmark
FActScore Leaderboard
FActScore evaluates factual precision in long-form text generation by assessing the support for atomic facts.
Updated Sep 5, 2026
Models2
Model coverage2
MetricScore
EvidenceB
FActScore Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelXA | Score97.03% | Percentile100.00% | Participants2 | EvidenceC | Evaluated |
| Rank02 | ModelOP | Score1.00% | Percentile0.00% | Participants2 | EvidenceC | Evaluated |
FActScore Highlights
The leading models and scores on this benchmark.
FActScore Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for FActScore
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is FActScore?
What FActScore measures and how its scores work.
FActScore is a fine-grained atomic evaluation metric for factual precision in long-form text generation.
It measures the percentage of atomic facts supported by reliable knowledge sources, with automated assessment using retrieval and language models.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- FActScore
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about FActScore.
Which model scores highest on FActScore?
Grok-4.1 is currently ranked first with 97.03%.
What are the top three models on FActScore?
The current leaders are Grok-4.1 (97.03%) and GPT-5 (1.00%).
Which FActScore model has the lowest official input price?
GPT-5 has the lowest matched official input price at $1.3 input / $10 output per 1M tokens.
Which models are fastest among FActScore results?
The fastest matched records are Grok-4.1 (100.00 tok/s via xAI) and GPT-5 (100.00 tok/s via OpenAI).
Does the highest score result prove overall model quality?
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.