Natural IM-Style Chat
Eidolon responds in a natural instant-messaging style, adapting tone, pacing, and expression to the current conversation context — less like issuing commands, more like talking to someone actually there.
Read the chat guide →Eidolon combines memory, tools, reasoning, and presence into one coherent digital self. Below is the full set of capabilities — from natural conversation to autonomous computer use.
Five capabilities that make Eidolon feel like a self — not a tool.
Eidolon responds in a natural instant-messaging style, adapting tone, pacing, and expression to the current conversation context — less like issuing commands, more like talking to someone actually there.
Read the chat guide →Low-latency voice conversations powered by veRTC with voiceprint noise reduction. Supports interruption handling, emotional tone, and long-running presence throughout the day.
Voice call docs →Face-to-face video presence for richer, more human interaction. Eidolon sees your environment through the camera and reacts in real time — a visual co-pilot for life and work.
Video call docs →Eidolon can create todos, build habits, generate images, set alarms, and operate tools on your behalf — through an async WebSocket pattern that never blocks the conversation.
Tool calling spec →A safer, more general-purpose automation layer than OpenClaw. Through secure remote interaction with terminal, browser, desktop screen, and keyboard/mouse, Eidolon can assist with gaming, document creation, vibe-coding orchestration, news summarization, stock analysis, smart-home control, and everyday computer operations.
Highest subscription tier (genesis) unlocks the deepest agent capabilities.
Agent architecture →Tools to help you build routines, track emotions, and stay focused — with AI by your side.
A unified activity feed with widgets and insights across all your life-tracking data — habits, mood, focus, todos, and journals in one place.
Capture tasks in natural language. Eidolon parses, organizes, and reminds you — and can break big tasks into smaller, scheduled steps.
Build good habits together. Streaks, gentle nudges, and AI-generated encouragement keep momentum without shame loops.
Track emotions and well-being over time. Eidolon spots patterns and offers grounded, non-clinical reflection prompts.
A Pomodoro-style deep-work timer with AI companionship. Silence distractions, log focus sessions, and review productivity trends.
Reminders that understand context. Eidolon can phrase, defer, and escalate alerts based on what you are doing and what matters.
Storytelling, image-making, expressive reactions, and integrations with the apps you already use.
Interactive, branching narrative experiences. Step into a character, make choices, and let Eidolon improvise the world around you.
Text-to-image creation inside the conversation. Generate avatars, illustrations, memes, and reference art without leaving Eidolon.
Expressive reactions and custom stickers. Eidolon can generate role-specific meme reactions in real time, tailored to the chat mood.
One-click conversation starters for common situations — morning check-in, focus session, wind-down, brainstorm, and more.
Eidolon can act as your AI presence on WeChat — auto-replying, summarizing messages, and helping you stay responsive without burnout.
Deploy Eidolon as a QQ channel bot. Bring companion presence and tool use directly into your QQ communities.
Tune the avatar, voice, theme, memory model, and capability tier to fit your life.
Bring your own VRM or Live2D model. Eidolon can also live in a floating desktop window so it stays present while you work.
9+ visual themes — aurora, bloom, dawn, diary, ember and more — plus font customization to match your aesthetic.
Customize Eidolon's voice or clone your own. Powered by Doubao S2S and FAL TTS for natural, expressive speech.
Eidolon maintains a persistent memory graph of people, projects, deadlines, preferences, and emotions — it never resets between conversations.
Extend Eidolon through the Model Context Protocol. Plug in new tools, data sources, and integrations without forking the core.
Four tiers — awakening, presence, perception, genesis — scale AI capability and quota. Default tier is configurable; genesis unlocks autonomous computer use.