One platform, every model
Ethyx plugs into external AI providers so you avoid vendor lock-in—without giving up a real workspace. History, projects, workflows, and security stay with you when you change models.
Closed beta with an access code today.
We encrypt chat history in our database; upstream AI providers still receive your prompts to generate responses.
Avoid vendor lock-in without a thin proxy
A model router that only forwards API calls still leaves you stranded when you want history, projects, MFA, search, images, and integrations in one place. Ethyx is the opposite: a full platform that happens to speak many providers. Pick the best model for the job—or let Auto route by task intent—while your conversations and workflows stay put.
Every major model, one workspace
OpenAI, Anthropic, xAI, DeepSeek, Kimi, Gemini, and Perplexity in the same client. Auto picks by task intent, or you choose explicitly—without juggling tabs and subscriptions.
A full platform, not a thin proxy
Encrypted chat history, projects, skills, compare mode, guided workflows, web search, and image generation. Switching models does not mean exporting your life to another app.
Honest about your data
Chat encrypted at rest, optional MFA, identity pseudonymized by default before provider requests—and a plain-language admission that upstream providers still receive prompts for inference.
Providers you can switch between
Provider breadth is convenience. The product wedge is the workspace you keep when you change models.
OpenAI
Anthropic
xAI
DeepSeek
Kimi
Gemini
Perplexity
Supported Providers
All the Major AI Models in One Place
OpenAI
Anthropic
xAI
DeepSeek
Kimi
Gemini
Perplexity