hcaptcha-challenger is a Python tool that aims to handle hCaptcha challenges using multimodal LLMs and integrated vision components. It provides a configurable workflow and supports multiple task types with pluggable resources.
Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).
What it is
hCaptcha Challenger. Gracefully face hCaptcha challenge with multimodal large language model. Documentation in multiple languages is linked in the repo: English, 简体中文, Русский, Tiếng Việt.
How it works
The project defines a set of challenge types and associates each with a pluggable resource and agent capability. It includes workflows for various tasks and references to contributors and model integrations. The README shows a table of supported challenge types with corresponding resources and capabilities, plus a workflow map for tasks and resources.
Getting started
The README includes setup and usage guidance across documentation links, plus references to models and pipelines. The Releases section lists version updates and changelog notes, indicating ongoing maintenance and feature additions. The latest release v0.19.0 (2026-01-05) and previous releases are listed with brief change notes.
Recent releases
- v0.19.0 (2026-01-05): Gemini 3 integration by @QIN2DIM in pull request #1082. Full changelog at the compare URL.
- v0.18.13 (2025-10-26): CaptchaTaskEntity/CaptchaTask fields optional with defaults by @QIN2DIM in pull request #1079.
- v0.18.12 (2025-10-22): Add retry logic and enhanced rule inference for spatial challenges by @QIN2DIM in pull request #1078.
- v0.18.11 (2025-10-16): MAX_CRUMB_COUNT config and improved crumb detection; added delay option by @QIN2DIM in PR #1074.
- v0.18.10 (2025-07-20): Enable debug mode by @QIN2DIM in PR #1057.
Traction
stars_7d and stars_1d are not provided in the Facts, so this section is omitted per the data. The repository has 2430 stars and 437 forks, with 48 open issues.
Behind the repo
The README notes integration with multiple external models and toolchains (Gemini, CLIP-ViT, YOLOv8 ONNX, ViT ONNX, etc.) and references to OpenAI, clip, and various AI-agent concepts. No explicit startup/company link is provided beyond the GitHub repository and the project’s “Contributors” section.
Caveats
- LICENSE: GPL-3.0
- Created: 2022-02-15
- Last push: 2026-01-28
- Language: Python
- License and open issues: 48 open issues
- Age and activity evidenced by releases through 2026-01-05 and ongoing maintenance