| license: mit | |
| size_categories: | |
| - 1K<n<10K | |
| # OmniAI OCR Benchmark | |
| A comprehensive benchmark that compares OCR and data extraction capabilities of different multimodal LLMs such as gpt-4o and gemini-2.0, evaluating both text and JSON extraction accuracy. | |
| [**Benchmark Results (Feb 2025)**](https://getomni.ai/ocr-benchmark) | [**Source Code**](https://github.com/getomni-ai/benchmark) |