Home/LLM Frameworks/heretic

heretic

p-e-w/heretic

Fully automatic censorship removal for language models

GitHub Website

GraphCanon updated Sep 20, 2026 · GitHub synced Sep 20, 2026

48views this month

32k stars3.5k forksLast push Sep 5, 2026 Python AGPL-3.0

Decision brief

Heretic is a Python-based tool for automatic censorship removal in language models, using transformer technology.

Good fit when

When you need to automatically remove censorship from language models and are comfortable with the AGPL-3.0 license, which requires that any derivative works be open source.
If your project involves working with transformer-based language models and you require a tool that can handle censorship removal without manual intervention.

Avoid when

If your project cannot comply with the AGPL-3.0 license, as it mandates that any derivative works must also be open source.
When you are looking for a tool that does not automatically remove censorship, as Heretic is specifically designed for this purpose and may not align with projects that require manual or selective de-

Requirements:: Heretic is a Python-based tool, so Python must be installed on the system where it will be used.

Observed Sep 18, 2026 · Source: enrich:decision_facts

heretic alternatives Trust report

Verify the decision

Maintenance and security

Full trust report

Maintenance: Active (12d since push); As of Sep 18, 2026
Provenance: Not a fork · Personal account; As of Sep 18, 2026
Security (OSV): No lockfile; As of Sep 18, 2026

Public GitHub metadata and optional OSV scans. Signals, not a guarantee. Trust methodology.

Install

pip install heretic

PyPI

How it fits your stack(5)

Typed graph edges - alternatives, integrations, successors, and dependencies. Ranked by relationship type, not raw GitHub stars.

Integrates

ray

Ray can integrate with Heretic for advanced usage scenarios that require the removal of censorship and scaling tasks involving sensitive text data.

integrates

Depends on

transformers

🤗 Transformers: state-of-the-art machine learning models for text, vision, audio, and multimodal models

depends on

Qwen

Heretic supports decensoring a variety of transformer-based language models, including Qwen. Therefore, it depends on Qwen for its operation if applied to Qwen instances.

depends on

Megatron-LM

Ongoing research training transformer models at scale

Qwen-VL

Official repo of Qwen-VL, a large vision language model by Alibaba Cloud

Relationship graph

Optional deeper exploration of typed edges and category neighbours.

Similar tools

Same-category neighbours not already linked as typed edges.

tokenizers

Fast State-of-the-Art Tokenizers optimized for Research and Production

11k★

presidio

A framework for detecting and anonymizing sensitive data

11k★

awesome-LLM-resources

Summary of the world's best LLM resources.

9.0k★

Evidence and technical details

Sourced facts, taxonomy, compatibility claims, README excerpt, and machine-readable endpoints.

Overview

A Python-based tool designed for automatic removal of censorship in language models, utilizing transformer technology.

Capability facts

CLI: CLI entrypoint
Source: pyproject.toml:[project.scripts] · Sep 18, 2026
Languages: python
Source: github.language+pyproject.toml · Sep 18, 2026

Graph entities

Smolagentsframework

README

License Copyright © 2025 2026 Philipp Emanuel Weidmann (<pew@worldwidemann.com ) + contributors This program is free software: you can redistribute it and/or modify it under the terms of the GNU Affero General Public License as published by the Free Software Foundation, eith...

For agents

This page has a .md twin and JSON over the API.

/tools/p-e-w-heretic.md GET /api/graphcanon/tools/p-e-w-heretic GET /api/graphcanon/graph?tool=p-e-w-heretic