NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
PyPI · #2873 most downloaded on PyPI
End-to-End Multi-Lingual Optical Character Recognition (OCR) Solution
Last release 2 years ago
no release in 18 months
Release timing varies
gaps range from 2 weeks to 13 months
Most releases are documented
notes for 22 of 33 stable releases
Nothing withdrawn
no release was ever pulled
6 years old
33 releases · first in 2020
One column per quarter.
24 September 2024 - Version 1.7.2
4 September 2023 - Version 1.7.1
Add Apple Silicon support (thanks @rayeesoft and @ArtemBernatskyy , see PR )
15 September 2022 - Version 1.6.2
1 September 2022 - Version 1.6.1
cyrillic_g2. This model is a new default for Cyrillic script.Restructure code to support alternative text detectors.
v1.6.0
DBNET, see paper. It can be used by initializing like this reader = easyocr.Reader(['en'], detect_network = 'dbnet18').Add trainer for CRAFT detection model (thanks @gmuffiness , see PR )
v1.5.0
Update dependencies (opencv and pillow issues)
11 September 2021 - Version 1.4.1
readtextlang method (thanks@arkya-art, see PR)rotation_info argument to support all possible angle (thanksabde0103, see PR)Instruction on training/using custom recognition model
Faster greedy decoder (thanks @samayala22)
Add support for PIL image (thanks @prays)
x_ths and y_ths to control merging behavior when paragraph=TrueNothing published for this version
Second-generation models: multiple times smaller size, multiple times faster inference, additional characters, comparable accuracy to the first genera
recog_network argument when creating Reader instance.
For example, reader = easyocr.Reader(['en','fr'], recog_network = 'latin_g1') will use the 1st generation Latin model.Nothing published for this version
22 February 2021 - Version 1.2.5
7 February 2021 - Version 1.2.4
1 February 2021 - Version 1.2.3
setLanguageList method to Reader class. This is a convenient api for changing languages (within the same model) after creating class instance.Add optimal_num_chars to detect method. If specified, bounding boxes with estimated number of characters near this value are returned first. (thanks @
optimal_num_chars to detect method. If specified, bounding boxes with estimated number of characters near this value are returned first. (thanks @adamfrees)rotation_info to readtext method. Allow EasyOCR to rotate each text box and return the one with the best confident score. Eligible values are 90, 180 and 270. For example, try [90, 180 ,270] for all possible text orientations. (thanks @mijoo308)Nothing published for this version
New language supports for Telugu and Kannada. These are experimental lite recognition models. Their file sizes are only around 7% of other models and
New language supports for Telugu and Kannada. These are experimental lite recognition models. Their file sizes are only around 7% of other models and they are ~6x faster at inference with CPU.
This release is also a preparation for user-created models/architectures in the future.
12 October 2020 - Version 1.1.10
beamsearch decoder (thanks @amitbcp)detect and recognize method for performing text detection and recognition separatelyNothing published for this version
20 new language supports for Bengali, Assamese, Abaza, Adyghe, Kabardian, Avar, Dargwa, Ingush, Chechen, Lak, Lezgian, Tabassaran, Bihari, Maithili, A
min_size argument for readtext: for filtering out small text boxTemporary fix for memory leakage on CPU mode
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Stable version with added simplified Chinese support
Stable version with added simplified Chinese support
Nothing published for this version
Your coding agent can read these notes before it upgrades. Set up the MCP server →