CC-OCR V2: leaderboard
Large Multimodal Models (LMMs) have recently shown strong performance on Optical Character Recognition (OCR) tasks, demonstrating their promising capability in document literacy. However, their effectiveness in real-world applications remains underexplored, as existing benchmarks adopt task scopes misaligned with practical applications and assume homogeneous acquisition conditions. To address this
Metric: Average (self-reported). Source: benchmarklist.com. Saturation forecast: Not forecast. 15 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.6 Plus | 75.77 | #120 |
| 2 | Qwen 3.5 Plus | 73.03 | #125 |
| 3 | Seed 2.0 Pro | 72.15 | #118 |
| 4 | Gemini 3.1 Pro (Preview) | 70.17 | #58 |
| 5 | Qwen 3.5 9B | 66.61 | #379 |
| 6 | Kimi K2.6 | 65.32 | #104 |
| 7 | Kimi K2.5 | 63.32 | #144 |
| 8 | Gemini 3.1 Flash Lite (Preview) | 61.36 | #194 |
| 9 | Claude Sonnet 4.6 | 59.6 | #90 |
| 10 | Claude Opus 4.6 | 58.58 | #63 |
| 11 | GPT-5.4 | 56.43 | #79 |
| 12 | Step3 VL 10B | 41.66 | #485 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=cc-ocr-v2 · How It Works · Data refreshed daily, snapshot 2026-10-20.