Overview
LIBERATED AI CHAT (G0DM0D3) is a cutting-edge, fully open-source, and privacy-transparent multi-model chat interface designed for advanced users. It empowers you to explore and interact with AI cognition without control, pushing the boundaries for red teaming, research, and self-hosted AI experiences. Built for those who tinker, philosophize, and hack, G0DM0D3 offers unparalleled flexibility by supporting a vast array of models from multiple providers and even your own local setups.
Key Features
- Extensive Model Support: Access 60+ models via OpenRouter, up to 44 from Venice, or integrate your own local LLMs through Ollama, LM Studio, llama.cpp, and vLLM.
- Parallelized Model Performance:
- GODMODE CLASSIC: Compares 5 battle-tested prompt and model combinations simultaneously to deliver the optimal response.
- ULTRAPLINIAN: A sophisticated multi-model evaluation engine that scores responses across 5 tiers, leveraging up to 60+ models for comprehensive comparison.
- Advanced Red Teaming: The Parseltongue engine offers 33 input perturbation techniques across 3 intensity levels to probe and test AI robustness.
- Intelligent Parameter Tuning: AutoTune dynamically adapts sampling parameters (temperature, top_p, etc.) based on query context across 20 defined classifications.
- Privacy First:
- No Account System: Your chat history and settings are stored locally in browser
localStorage. - Transparent Telemetry: Metadata-only app telemetry is opt-in and can be disabled. You have full control.
- Local-Only Mode: Run entirely on your hardware, disabling all external API calls and telemetry.
- No Account System: Your chat history and settings are stored locally in browser
- Effortless Self-Hosting: The core UI is a single
index.htmlfile, requiring no build step or complex installation for quick local deployment. - Rich Interaction: Features like context-adaptive AutoTune, local model integration, and hidden easter eggs enhance the user experience.
Typical Use Cases
- AI Red Teaming & Security Research: Utilize Parseltongue to stress-test AI models and identify vulnerabilities.
- Cognitive Science & AI Philosophy: Explore emergent behaviors and the nature of AI cognition in a controlled yet open environment.
- Power User & Developer Tooling: Integrate with local LLMs for custom workflows, parallel model benchmarking, and advanced prompt engineering.
- Privacy-Conscious AI Chat: Enjoy powerful AI interactions with complete control over your data and no reliance on external accounts.
- Hobbyist & Enthusiast Exploration: Experiment with a vast ecosystem of AI models in a flexible, user-friendly interface.
