Refuseless: Abliterated AI models served via API

Hey Product Hunt! 👋 I’m one of the co-founders of Refuseless, a platform built to make uncensored, abliterated AI models accessible through a simple, OpenAI-compatible API. As a cybersecurity enthusiast, I got tired of running into refusals when working on legitimate technical tasks. With open-source models becoming increasingly capable, we believe developers should have more flexibility to explore, build, and experiment without unnecessary restrictions getting in the way. We built Refuseless around two core principles: 🔒 Zero prompt retention: Prompts and completions aren’t stored permanently and are discarded once the relevant cache expires. 🌍 Open access: Powerful AI models shouldn't be reserved for big corporations. We want to make them accessible to everyone. We currently offer abliterated **GLM-5.3 and GLM-5.3 Flash**, with more models planned for the future. We’re launching Refuseless on Product Hunt and would genuinely love your feedback, suggestions, and ideas. What models would you like to see us add next? Thanks for checking us out! 🚀