Alla program > Filmproduktion > Vtuber Assistant
Gemenskapscentral
Vtuber Assistant
Your AI co-host desktop companion & beginner VTuber rig in one. Toggle live facial tracking or let AI react to screen, audio, Twitch chat with hype commentary. Drag-resize overlay for streams, layer animations, custom prompts—local processing, pluggable models, OBS-ready. BEGINNER FRIENDLY!!!

Registrera dig för att lägga till denna artikel i din önskelista eller ignorera den.

Funktioner
Steam Cloud
Steam håller på att lära sig mer om detta program 
Kräver godkännande av EULA hos tredje part
Språk:
Gränssnitt Fullt ljud Undertexter
Svenska Stöds inte
Engelska ✔ ✔
Innehåll
Inkluderar interaktiva element

Interaktivitet online

Early Access-program

Bli involverad i det här programmet medan det utvecklas.

Notera: Detta Early Access-program är inte färdigt och skulle kunna ändras ytterligare. Om du inte känner för att använda det i sin nuvarande form bör du vänta för att se om det fortsätter utvecklas mer. Läs mer

Vad utvecklarna har att säga:

Varför Early Access?

“Vtuber Assistant is a passion project born from wanting a real AI companion that could both act as a desktop pet and double as a beginner-friendly VTuber rig. Early Access lets me involve the community early—your feedback on AI reactivity, tracking polish, pluggable integrations, animation layering, and overall usability will directly shape updates. This is the best way to build something that truly fits what streamers, VTubers, and casual users actually want.”

Ungefär hur länge kommer det här programmet att vara i Early Access?

“We plan for Early Access to last 6–12 months, depending on community feedback and feature pace. The goal is to reach a stable 1.0 with core polish, expanded reactivity, and robust pluggable support, but the timeline will flex based on what matters most to users.”

Hur är det tänkt att den fullständiga versionen ska skilja sig från versionen i Early Access?

“We plan to expand scope in these areas during Early Access and beyond:

-Deeper AI brain integration (more contextual synthesis, better multi-input understanding).
-Expanded reaction library (pre-made & user-contributed animations, expressions, VFX).
-More pluggable options (additional model support, advanced TTS/STT integrations).
-UI/UX polish (hotkey customization, preset saves, better onboarding for beginners).
-Performance & stability improvements across Windows, Linux, macOS, and Steam Deck.
-We won't make hard promises on exact timelines or features—development will evolve with your input.”

Hur är det nuvarande läget i versionen med Early Access?

“The Early Access version is fully playable and feature-complete for core use cases:

-Drag-and-drop desktop overlay with resize/positioning (OBS-friendly).
-Toggle between user-controlled facial tracking (webcam, overlay mask, extra bones, sliders/toggles) and AI co-host mode.
-Live reactivity to system audio, visual screen capture, Twitch chat, and voice/text input.
Generative AI responses (local default + user-pluggable online), TTS voice output, layered animations under tracking.
-Custom personality prompts, safety guard, and modular pluggable system.
It's stable for daily use (streaming, gaming, casual hanging out), but will continue to receive polish, bug fixes, and new features based on feedback.”

Kommer det här programmet att prissättas annorlunda under och efter Early Access?

“I want to be clear, I do not plan to increase the price after early access. If anything I have thoughts to lower it depending on how wide spread this goes. To be honest I am playing this by ear.”

Hur planerar ni att involvera gemenskapen i er utvecklingsprocess?

“Community input is central:

-Steam reviews, discussions, and feedback threads will guide priorities (e.g., which integrations or reactions matter most).
-In-app feedback forms and Discord (if created) for direct suggestions.
Regular devlogs/news updates on Steam showing what's being worked on and why.
-Beta branches for testing new features before wider release.
-Roadmap visibility: I'll share planned additions and adjust based on what users vote for or request most.”
Läs mer
Det här spelet är ännu inte tillgängligt på Steam

Planerat lanseringsdatum: 24 apr, 2026

Det här programmet kommer att låsas upp om 6 veckor

Intresserad?
Lägg till på din önskelista och bli meddelad när det släpps.
 
Se alla diskussioner

Rapportera buggar och lämna feedback för det här programmet på forumet

EARLY ACCESS LAUNCH!!! – Shape the Future!

Vtuber Assistant is soon to be in Early Access! Get in early, and help guide development. Roadmap highlights: smarter AI brain (aggregate from Visual/Audio/User input), more animation layering, expanded pluggable support, reaction library, and your feedback directly shapes updates. Join the squad—wishlists & reviews help a ton!

Om detta program

Vtuber Assistant is your all-in-one AI-powered desktop companion and beginner-friendly VTuber rig — built because I wanted a real Jarvis-style buddy that watches, reacts, and hypes without the fluff.

Toggle between two modes anytime:

AI Control Let your customizable AI co-host take over — reacting to your gameplay screen, system audio, Twitch chat, and voice input with thoughtful commentary, animations, and voice lines. It sees the big picture: your blindfolded 360 no-scope, chat going wild, bass drop in the music — and drops the perfect hype line.

User Control Plug in your webcam (C920 works great!) and drive the avatar live with facial tracking, overlay mask, extra bones (spine rotations, facial details), sensitivity sliders, toggles, and layered animations that play under your real expressions — no desync, just silky motion.

Key Features • Drag-and-drop desktop overlay: Resize, position anywhere — perfect for OBS corner cam or reaction shots. • Pluggable everything: Bring your own AI models, TTS/STT, animations, personality prompts — defaults work instantly. • Screen-aware brain: AI watches selected screen visuals + audio, Twitch chat, your voice — generates contextual reactions every few minutes (editable prompt, frequency slider). • Layered animations: Trigger victory spins, dances, fist pumps on top of live tracking — hotkey support coming soon. • Local-first & privacy-focused: All core processing on-device by default (bundled Ollama LLM + Piper TTS). External services optional via your own API keys. • Beginner to pro: Start simple as a cute pet, flip to full VTuber mode when ready — no watermarks, no subscriptions.

Whether you're grinding ranked, streaming, making content, or just want a dance partner while working, Vtuber Assistant grows with you. Early Access means your feedback shapes the future — join the squad and help build the ultimate AI sidekick!

No art was harmed by AI in the production of this software. Everything you see is hand-crafted vibes and your own creativity.

Twitch Chat functionality is read-only by the AI and produces in-app responses only.

Upplysning om AI-genererat innehåll

Utvecklarna beskriver hur deras spel använder AI-genererat innehåll så här:

Vtuber Assistant uses generative AI to create live text responses and voice output as a customizable co-host companion. It reacts to your voice/text input, screen audio/visuals, and Twitch chat with fun, contextual commentary.
All core functionality runs locally by default (bundled Ollama LLM + Piper TTS) — no external setup or payment required to use the app. Online AI/TTS/STT services (e.g., Grok, ElevenLabs, OpenAI-compatible APIs) are completely optional and fully user-configured (bring your own API keys).
Twitch chat integration is read-only — the AI consumes incoming messages and produces in-app responses only.
All outputs are moderated by a built-in Safety Guard to help prevent inappropriate content. No in-app purchases, subscriptions, or DLC for AI features.

Systemkrav

Windows
SteamOS + Linux
    Minimum:
    • OS: Windows 10 (64-bit) or later
    • Processor: Intel Core i5 6th Gen / AMD Ryzen 3 or equivalent
    • Minne: 8 GB RAM
    • Grafik: Integrated GPU with WebGL 2.0 support (Intel UHD 630 / AMD Radeon RX 550 or better)
    • Lagring: 2 GB ledigt utrymme
    • Ljudkort: Any
    • Ytterligare anmärkningar: Webcam (720p+) required for face tracking; Microphone for voice input. Works fully offline with bundled AI.
    Rekommenderade:
    • OS: Windows 11 (64-bit)
    • Processor: Intel Core i7 9th Gen+ / AMD Ryzen 5 5600X+ or better
    • Minne: 16 GB RAM
    • Grafik: Dedicated GPU (NVIDIA GTX 1660 / AMD RX 5600 XT or equivalent)
    • DirectX: Version 12
    • Nätverk: Bredbandsanslutning
    • Lagring: 6 GB ledigt utrymme
    • Ljudkort: Any
    • Ytterligare anmärkningar: Dedicated GPU accelerates vision AI models. 1080p webcam recommended.
    Minimum:
    • OS: Ubuntu 20.04+ / Fedora 38+ / SteamOS 3+ or equivalent modern distro
    • Processor: Intel Core i5 6th Gen / AMD Ryzen 3 or equivalent
    • Minne: 8 GB RAM
    • Grafik: GPU with WebGL 2.0 support (Intel/AMD integrated or NVIDIA/AMD dedicated)
    • Lagring: 2 GB ledigt utrymme
    • Ljudkort: Any
    • Ytterligare anmärkningar: Webcam (720p+) and microphone required. Tested on Bazzite/Steam Deck. Flatpak or AppImage recommended for immutable distros.
    Rekommenderade:
    • OS: Latest Ubuntu / Fedora / SteamOS
    • Processor: Intel Core i7 9th Gen+ / AMD Ryzen 5 5600X+
    • Minne: 16 GB RAM
    • Grafik: Dedicated GPU (NVIDIA/AMD with ROCm/CUDA support preferred)
    • Nätverk: Bredbandsanslutning
    • Lagring: 6 GB ledigt utrymme
    • Ljudkort: Any
    • Ytterligare anmärkningar: GPU acceleration for AI models; Steam Deck compatible in desktop mode.
Det finns inga recensioner för denna produkt

Du kan skriva din egen recension för den här produkten för att dela din upplevelse med gemenskapen. Använd området ovanför köpknapparna på den här sidan för att skriva din recension.