CC-OCR
We show this table for reference; we do not rank on it.
An OCR-focused benchmark for reading and extracting text from visually complex documents and images.
Benchmark score on CC-OCR — November 13, 2026
We compile the CC-OCR rows from provider self-reports and secondary reports. Qwen3.5 397B leads the table at 82.0%, followed by Qwen3.6-35B-A3B (81.9%) and Qwen3.6-27B (81.2%). We do not use these results to rank models overall.
Qwen3.5 397B
Alibaba
Reported by AlibabaQwen3.6-35B-A3B
Alibaba
Reported by AlibabaQwen3.6-27B
Alibaba
Reported by Alibaba7 modelsMultimodal & GroundedCurrentDisplay onlyUpdated November 13, 2026
Benchmark score table (7 models)
None of the 7 results on this page has a documented run by the benchmark or an independent evaluator: 4 reported by the labs that built the models and 3 reported in another publisher’s document.
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | Qwen3.5 397BAlibabaReported by Alibaba | 82.0% | Reported by Alibaba | Not reported | Open |
| 2 | Qwen3.6-35B-A3BAlibabaReported by Alibaba | 81.9% | Reported by Alibaba | Not reported | Open |
| 3 | Qwen3.6-27BAlibabaReported by Alibaba | 81.2% | Reported by Alibaba | Not reported | Open |
| 4 | Kimi K2.5Moonshot AIAnother lab’s comparison table | 79.7% | Another lab’s comparison table | Not reported | Open |
| 5 | Qwen3.8 MaxAlibabaReported by Alibaba | 79.6% | Reported by Alibaba | Not reported | Open |
| 6 | Gemini 3 ProGoogleAnother lab’s comparison table | 79.0% | Another lab’s comparison table | Not reported | Closed |
| 7 | GPT-5.2OpenAIAnother lab’s comparison table | 70.3% | Another lab’s comparison table | Not reported | Closed |
- Lab-reported
- Benchmark-run
- Independent
Among the reported CC-OCR rows, Qwen3.5 397B is first at 82.0%. The third row is 0.8 points behind. The broader top-10 range is 11.7 points, so the table still separates the published systems.
7 models have been evaluated on CC-OCR. The benchmark falls in the Multimodal & Grounded category. CC-OCR is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About CC-OCR
Year
2026
Tasks
Optical character recognition
Format
Text extraction from images and documents
Difficulty
Document reading
CC-OCR is useful as a direct check on raw reading ability before higher-level reasoning. It highlights whether failures come from extraction quality or from later reasoning over the extracted content.
Freshness and provenance
Version
CC-OCR 2026
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
What does CC-OCR measure?
An OCR-focused benchmark for reading and extracting text from visually complex documents and images.
Which model scores highest on CC-OCR?
Qwen3.5 397B by Alibaba currently leads with a score of 82.0% on CC-OCR.
How many models are evaluated on CC-OCR?
7 AI models have published results on CC-OCR in the BenchLM catalog.
Compare top models on CC-OCR
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 7,000+ subscribers.
One email each week. Unsubscribe anytime.