keras-ocr
A packaged and flexible version of the CRAFT text detector and Keras CRNN recognition model.
Decision gist · record as of 2026-08-14
Yes, with conditions. The package is stable, has no known vulnerabilities, and low install friction. It's suitable for projects that need local OCR without cloud dependencies. However, the aging maintenance status (last release 1012 days ago) means it may not track the latest TensorFlow or dependency versions—verify compatibility with your environment before committing to production use. Best for new projects where you can pin dependencies, or for prototyping where you can tolerate potential future compatibility issues.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires TensorFlow >= 2.0.0 and Python >= 3.7.5.
- Pre-trained model weights are downloaded automatically on first use.
- GPU memory management can be configured via MEMORY_GROWTH and MEMORY_ALLOCATED environment variables.
License · maintenance · safety
MIT (permissive) — MIT license is permissive, allowing commercial and private use with minimal restrictions. You may use, modify, and distribute the package freely as long as you include the original license notice.
last release 2023-11-06 (1012 days) · last repo commit 2025-09-22 · 1,474 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 108,017 downloads/mo, #12,584 on PyPI
Alternatives
Verify before relying
pip install keras-ocr
import keras_ocr
pipeline = keras_ocr.pipeline.Pipeline()
images = [keras_ocr.tools.read(url) for url in ['https://example.com/image.jpg']]
predictions = pipeline.recognize(images)- Whether the package works reliably with TensorFlow 2.x versions released after November 2023.
- Current performance on modern OCR benchmarks or whether the published metrics remain representative.
- Compatibility with recent versions of imgaug and other dependencies given the aging maintenance status.
What it is and what it does
keras-ocr is a packaged implementation of two published deep learning models: the CRAFT text detector (from Clova AI) and the CRNN recognition model. It wraps these into a single pipeline that takes images as input and returns detected text regions with recognized words. The package ships with pre-trained weights and downloads them automatically, so you can start extracting text from images with minimal setup.
The library is designed for scene text (text that appears naturally in photographs) rather than scanned documents. It exposes a straightforward API—create a pipeline, pass images to it, and receive (word, bounding-box) tuples. The fact sheet shows it performs comparably to cloud OCR services on the COCO-Text benchmark, though it ignores punctuation and case by design. It depends on TensorFlow, image processing libraries (imgaug, shapely, fonttools), and utility packages (editdistance, validators, tqdm).
Use it for
- Extract text from photographs of signs, labels, or documents for indexing or data entry automation.
- Build a local OCR pipeline that doesn't require cloud API calls or internet connectivity.
- Train a custom text detection or recognition model using the provided framework and your own labeled data.
- Batch-process images to extract and validate text content as part of a data pipeline.
- Integrate text detection into computer vision applications that need to locate and read text regions.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, with conditions.
The package is stable, has no known vulnerabilities, and low install friction. It's suitable for projects that need local OCR without cloud dependencies. However, the aging maintenance status (last release 1012 days ago) means it may not track the latest TensorFlow or dependency versions—verify compatibility with your environment before committing to production use. Best for new projects where you can pin dependencies, or for prototyping where you can tolerate potential future compatibility issues.
Install
keras-ocr on PyPI
Before you install
Low install friction with a pure-Python wheel. Maintenance status is aging—last release was 1012 days ago (November 2023) and the last commit was September 2025, suggesting the project is maintained but not actively developed. The 9 runtime dependencies are standard ML/image processing libraries.
Requires TensorFlow >= 2.0.0 and Python >= 3.7.5. Pre-trained model weights are downloaded automatically on first use. GPU memory management can be configured via MEMORY_GROWTH and MEMORY_ALLOCATED environment variables.
License in practice
MIT license is permissive, allowing commercial and private use with minimal restrictions. You may use, modify, and distribute the package freely as long as you include the original license notice.
Quickstart
pip install keras-ocr
import keras_ocr
pipeline = keras_ocr.pipeline.Pipeline()
images = [keras_ocr.tools.read(url) for url in ['https://example.com/image.jpg']]
predictions = pipeline.recognize(images)Verify before relying
- Whether the package works reliably with TensorFlow 2.x versions released after November 2023.
- Current performance on modern OCR benchmarks or whether the published metrics remain representative.
- Compatibility with recent versions of imgaug and other dependencies given the aging maintenance status.
Package facts
| License | MIT permissive |
| Python support | Supports the current Python release >=3.7.5 |
| Install friction | Low. Pure-Python wheel |
| Runtime dependencies | 9 packageseditdistanceefficientnetessential_generatorsfonttoolsimgaugpyclippershapelytqdmvalidators |
| Maintenance | Aging 1,012 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 108,017 / month, #12,584 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | License :: OSI Approved :: MIT LicenseOperating System :: MacOSOperating System :: Microsoft :: WindowsOperating System :: POSIX :: LinuxOperating System :: UnixProgramming Language :: Python :: 3Programming Language :: Python :: 3 :: OnlyProgramming Language :: Python :: 3.10Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.8Programming Language :: Python :: 3.9 |
Evidence: keras_ocr-0.9.3-py3-none-any.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “CRAFT CRNN pipeline”
- keras-ocrkeras-ocr detects and recognizes text in images using pre-trained…
- craft-partsCraft Parts provides a declarative framework for obtaining,…
- python-doctrpython-doctr extracts and localizes text from documents and images…
Give your agent the search over MCP, or paste the wish link into any chat.
More Artificial Intelligence packages
LiteLLM provides a unified Python interface to call 100+ LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, and others) using OpenAI-compatible API format, available as both a Python SDK and a self-hosted AI Gateway proxy server.
Install it if you need to work with multiple LLM providers or want to centralize LLM routing in your organization.
Client library and CLI tool for downloading, uploading, and managing models, datasets, and repositories on the Hugging Face Hub platform.
Install it if you work with Hugging Face Hub models or datasets.
LangChain provides a framework for building agents and LLM-powered applications by composing language models, tools, and memory through a unified API that abstracts over multiple model providers.
hf-xet provides chunk-based deduplication and efficient file transfer for the Hugging Face Hub, enabling faster uploads and downloads of large files with local disk caching.
Tokenizers converts raw text into token sequences for NLP models, with support for training custom vocabularies and using pre-built tokenizers (BPE, WordPiece) optimized for speed via Rust.
Transformers provides a unified framework for loading, fine-tuning, and running state-of-the-art pretrained models across text, vision, audio, video, and multimodal tasks using PyTorch, JAX, or TensorFlow.
Install it if you need to run or train any transformer-based model for NLP, vision, audio, or multimodal tasks.
See also easyocr · python-doctr · tesserocr · unstructured.pytesseract · pytesseract · pyocr · surya-ocr · onnxtr · keras-nlp · ocrmac