Hi Product Hunt! I'm excited to introduce Abolitus. As AI becomes more integrated into our lives, privacy and freedom of expression are often compromised by platforms that log data and police prompts. We built Abolitus because we believe your AI conversations—especially roleplay—should be completely private and yours alone. Abolitus takes a radically different approach: The entire chat engine runs client-side in your browser. Here is what makes Abolitus unique: Zero-Knowledge Architecture: Our backend acts purely as a dumb, encrypted blob store. Your master encryption key never leaves your browser. We have zero knowledge of your data, and we keep zero logs. PC-to-Mobile E2EE Tunneling: Run a local LLM (like Ollama or LM Studio) on your home PC, and seamlessly interact with it from your phone anywhere in the world. We tunnel the request through our central server via strict End-to-End Encryption. No port forwarding, same-network requirements, or complicated networking setups needed. Cross-Platform & E2EE Cloud Sync: Abolitus is a web-based client that works seamlessly across mobile, PC, and tablets. Your chats, characters, and settings are synced across all your devices using E2EE, meaning your data is always accessible to you, but mathematically impossible for us to read. Multi-API Hydra: Bring your own model. Connect directly to OpenRouter, NanoGPT or run local models (Ollama/LM Studio) on your localhost. There is no API middleman monitoring your requests. 100% Local Processing: Heavy computation, including vector embeddings and RAG (Retrieval-Augmented Generation) via Transformers.js, runs completely locally on your device. If you value privacy, technical transparency, and an uncensored AI experience, Abolitus was built for you. I'd love for you to try it out, plug in your preferred models, and let me know your thoughts or questions below!

Abolitus

Privacy-first, uncensored AI roleplay in your browser.

Privacy-first, uncensored AI roleplay in your browser.