Benchmark
OCRBench_V2
OCR benchmarks measure how well models read text in images and documents. Scores are percentages. Higher is better.
- Models measured
- 10
- Scores recorded
- 11
- Sources
- 2
- IQ weight
- 2.35
Score distribution
- 61.4%
- Mean
- 62.8%
- Median
- 3.7%
- Std dev
- 56.1%
- 10th percentile
- 66.3%
- 90th percentile
- 56.1%
- Min
This benchmark has a weight of 2.35 for chat models, measured over 9 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | Qwen3.7 Plus· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 67.1% |
| 2 | Kimi K2.5SenseNova 6.7 Flash-Lite official benchmark chart · chart setting· Reported by SenseTime / OpenSenseNova· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 66.3% |
| 3 | GPT 5.4SenseNova 6.7 Flash-Lite official benchmark chart · chart setting· Reported by SenseTime / OpenSenseNova· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 64.1% |
| 4 | Doubao Seed 2.1 Pro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 63.2% |
| 5 | Gemini 3.1 Pro PreviewSenseNova 6.7 Flash-Lite official benchmark chart · chart setting· Reported by SenseTime / OpenSenseNova· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 63% |
| 6 | Doubao Seed 2.1 Turbo· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 62.8% |
| 7 | Claude Opus 4.6SenseNova 6.7 Flash-Lite official benchmark chart · chart setting· Reported by SenseTime / OpenSenseNova· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 60.3% |
| 8 | Seed 2.0 Mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 58.5% |
| 9 | Qwen2.5-Omni 7B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 57.8% |
| 10 | Nova 2 Lite· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 56.1% |
| 11 | Nova 2 LiteAmazon Nova 2 Lite official launch chart / technical report · average accuracy, v1· Reported by Amazon Web Services, Amazon AGI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 56.1% |
Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links. · Data by LLM Stats