multimodal benchmark
CC-OCR Leaderboard
CC-OCR is an OCR benchmark for evaluating Large Multimodal Models (LMMs) in literacy across four OCR-centric tracks.
Updated Sep 5, 2026
Models18
Model coverage18
MetricScore
EvidenceB
CC-OCR Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAC | Score83.40% | Percentile100.00% | Participants18 | EvidenceC | Evaluated |
| Rank02 | ModelAC | Score82.20% | Percentile94.12% | Participants18 | EvidenceC | Evaluated |
| Rank03 | ModelAC | Score81.90% | Percentile88.24% | Participants18 | EvidenceC | Evaluated |
| Rank04 | ModelAC | Score81.80% | Percentile82.35% | Participants18 | EvidenceC | Evaluated |
| Rank05 | ModelAC | Score81.50% | Percentile76.47% | Participants18 | EvidenceC | Evaluated |
| Rank06 | ModelAC | Score81.20% | Percentile70.59% | Participants18 | EvidenceC | Evaluated |
| Rank07 | ModelAC | Score81.00% | Percentile64.71% | Participants18 | EvidenceC | Evaluated |
| Rank08 | ModelAC | Score80.70% | Percentile58.82% | Participants18 | EvidenceC | Evaluated |
| Rank09 | ModelAC | Score80.70% | Percentile52.94% | Participants18 | EvidenceC | Evaluated |
| Rank10 | ModelAC | Score80.30% | Percentile47.06% | Participants18 | EvidenceC | Evaluated |
| Rank11 | ModelAC | Score79.90% | Percentile41.18% | Participants18 | EvidenceC | Evaluated |
| Rank12 | ModelAC | Score79.80% | Percentile35.29% | Participants18 | EvidenceC | Evaluated |
| Rank13 | ModelAC | Score77.80% | Percentile29.41% | Participants18 | EvidenceC | Evaluated |
| Rank14 | ModelAC | Score77.80% | Percentile23.53% | Participants18 | EvidenceC | Evaluated |
| Rank15 | ModelAC | Score77.10% | Percentile17.65% | Participants18 | EvidenceC | Evaluated |
| Rank16 | ModelAC | Score76.30% | Percentile11.76% | Participants18 | EvidenceC | Evaluated |
| Rank17 | ModelAC | Score76.20% | Percentile5.88% | Participants18 | EvidenceC | Evaluated |
| Rank18 | ModelAC | Score73.80% | Percentile0.00% | Participants18 | EvidenceC | Evaluated |
CC-OCR Highlights
The leading models and scores on this benchmark.
CC-OCR Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for CC-OCR
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is CC-OCR?
What CC-OCR measures and how its scores work.
CC-OCR comprises multi-scene text reading, multilingual text reading, document parsing, and key information extraction tracks, with 39 subsets and 7,058 fully annotated images.
It measures text grounding, multi-orientation text recognition, and hallucination/repetition detection across diverse visual challenges, reported as a Score ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- CC-OCR
- Modality
- multimodal
- Primary category
- multimodal
- 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 CC-OCR.
Which model scores highest on CC-OCR?
Qwen3.6 Plus is currently ranked first with 83.40%.
What are the top three models on CC-OCR?
The current leaders are Qwen3.6 Plus (83.40%), Qwen3 VL 235B A22B Instruct (82.20%), and Qwen3.6-35B-A3B (81.90%).
Which CC-OCR model has the lowest official input price?
Qwen3.6-35B-A3B has the lowest matched official input price at $0.25 input / $1.5 output per 1M tokens.
Which models are fastest among CC-OCR results?
The fastest matched records are Qwen3 VL 4B Thinking (8.77 tok/s via DeepInfra).
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.