image to text benchmark
OCRBench_V2 Leaderboard
OCRBench_V2 is a bilingual benchmark for evaluating Large Multimodal Models on visual text localization and reasoning.
Updated Sep 5, 2026
Models7
Model coverage7
MetricScore
EvidenceB
OCRBench_V2 Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAC | Score67.10% | Percentile100.00% | Participants7 | EvidenceC | Evaluated |
| Rank02 | ModelAM Nova 2 ProAmazon | Score64.50% | Percentile83.33% | Participants7 | EvidenceC | Evaluated |
| Rank03 | ModelBY | Score63.20% | Percentile66.67% | Participants7 | EvidenceC | Evaluated |
| Rank04 | ModelBY | Score62.80% | Percentile50.00% | Participants7 | EvidenceC | Evaluated |
| Rank05 | ModelAMNova 2 OmniAmazon | Score58.20% | Percentile33.33% | Participants7 | EvidenceC | Evaluated |
| Rank06 | ModelAC | Score57.80% | Percentile16.67% | Participants7 | EvidenceC | Evaluated |
| Rank07 | ModelAMNova 2 LiteAmazon | Score56.10% | Percentile0.00% | Participants7 | EvidenceC | Evaluated |
OCRBench_V2 Highlights
The leading models and scores on this benchmark.
OCRBench_V2 Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for OCRBench_V2
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is OCRBench_V2?
What OCRBench_V2 measures and how its scores work.
OCRBench_V2 contains 10,000 human-verified question-answering pairs across 8 core OCR capabilities.
It measures visual text localization and reasoning, reported as a Score ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- OCRBench_V2
- Modality
- multimodal
- Primary category
- image to text
- 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 OCRBench_V2.
Which model scores highest on OCRBench_V2?
Qwen3.7-Plus is currently ranked first with 67.10%.
What are the top three models on OCRBench_V2?
The current leaders are Qwen3.7-Plus (67.10%), Nova 2 Pro (64.50%), and Seed 2.1 Pro (63.20%).
Which OCRBench_V2 model has the lowest official input price?
Qwen2.5-Omni-7B has the lowest matched official input price at $0.10 input / $0.40 output per 1M tokens.
Which models are fastest among OCRBench_V2 results?
No matched runtime record is currently available.
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.