Steam release scheduled · App 4180890

AI Digital Human Lab

View on Steam ↗Updated Awaiting first sample
  • Single-player
  • Family Sharing
Steam

Store data

Store status
Coming soon
Scheduled release
2 November 2026
Positive reviews
—
Total reviews
—
Release date
2 November 2026
Player tracking—Begins when the game launches
Release date2 November 2026Scheduled by Steam
Store statusComing soonSteam store listing
Steam rating—0 reviews
Steam metadata

Game information

Details supplied by the Steam store
App ID
4180890
App type
Game
Supported systems
Windows · macOS · Linux
Genres
Casual, Indie, Simulation
Release date
2 November 2026
Last player update
Awaiting first sample
Steam store data

System requirements

Supplied by the developer on Steam

Windows

Minimum
Requirement
Requires a 64-bit processor and operating system
OS
Windows 10 64-bit
Processor
Intel Core i5-6500 / AMD Ryzen 5 1600
Memory
8 GB RAM
Graphics
NVIDIA GeForce GTX 1060 / AMD Radeon RX 580
DirectX
Version 12
Storage
20 GB available space
Sound Card
DirectX compatible sound card
VR Support
Not required
Additional Notes
Local AI models run fully offline — internet connection required only for optional cloud agents
Recommended
Requirement
Requires a 64-bit processor and operating system
OS
Windows 10 64-bit / Windows 11
Processor
Intel Core i7-9700 / AMD Ryzen 7 3700X
Memory
16 GB RAM
Graphics
NVIDIA GeForce RTX 3060 / AMD Radeon RX 6700 XT
DirectX
Version 12
Storage
25 GB available space
Sound Card
DirectX compatible sound card
VR Support
Not required
Additional Notes
A faster CPU improves local AI response speed; SSD recommended

macOS

Minimum
Requirement
Requires an Apple processor
OS
macOS 14 Sonoma or later
Processor
Apple M1 or later (Apple silicon only; Intel Macs are not supported)
Memory
16 GB RAM
Graphics
Apple silicon GPU (Metal 3)
Storage
20 GB available space
Sound Card
Built-in audio; a microphone for voice chat
Additional Notes
Local AI models run fully offline on Apple silicon — internet connection required only for optional cloud agents
Recommended
Requirement
Requires an Apple processor
OS
macOS 15 Sequoia or later
Processor
Apple M2 Pro / M3 / M4 or newer
Memory
24 GB RAM
Graphics
Apple silicon GPU (M2 Pro or newer)
Storage
25 GB available space
Sound Card
Built-in audio
Additional Notes
More unified memory and a faster chip speed up local AI replies and voice

Linux

Minimum
OS
SteamOS 3.6 or later (Steam Deck) / 64-bit Linux, e.g. Ubuntu 22.04 or newer
Processor
Steam Deck APU / Intel Core i5-6500 / AMD Ryzen 5 1600 (x86-64 with AVX2)
Memory
8 GB RAM
Graphics
Steam Deck / Vulkan 1.3 GPU: NVIDIA GeForce GTX 1060 / AMD Radeon RX 580 (Mesa 23+ or NVIDIA 535+)
Storage
20 GB available space
Sound Card
PipeWire or PulseAudio; a microphone for voice chat
VR Support
Not required
Additional Notes
Native Vulkan build, Steam Deck supported. Local AI models run fully offline — internet connection required only for optional cloud agents
Recommended
OS
SteamOS 3.7 or later / Ubuntu 24.04 64-bit or newer
Processor
Intel Core i7-9700 / AMD Ryzen 7 3700X
Memory
16 GB RAM
Graphics
NVIDIA GeForce RTX 3060 / AMD Radeon RX 6700 XT (Vulkan 1.3)
Storage
25 GB available space
Sound Card
PipeWire or PulseAudio
VR Support
Not required
Additional Notes
A faster CPU improves local AI response speed; SSD recommended
Explore charts ↓
Player history

Concurrent players

This range will fill as SteamWatch collects more samples.

24-hour and 7-day views use hourly peaks from ten-minute samples. Longer views use retained daily peaks. Up to 24 recent event markers are shown per layer.

Twitch

Twitch stats

Awaiting the first Twitch sample.
Long-term history

Monthly players breakdown

Calculated from SteamWatch samples
Monthly data will appear after the first successful player sample.
Explore game activity ↓
About this game

Description

Supplied by the Steam store

A Digital Human Who Listens, Understands, and Answers in Her Own Voice

Meet Luna, a lifelike digital human who lives right on your PC. Talk to her by voice or text — ask a question, share your day, tell a story — and she answers in the moment, in her own voice, with accurate lip-sync, 74 facial expressions and gestures that fit what she says. No scripts, no dialogue trees: every reply is generated live. Her language model and her voice run on your own graphics card, so your conversations stay on your PC — no subscription, no API keys.

Bring Your Own AI Agent — Built-in MCP

AI Digital Human Lab has a built-in MCP server (Model Context Protocol, the open standard AI assistants use to work with apps). Connect the AI agent you already use — Muse, Claude, or any other MCP-compatible assistant — and it can see the scene and direct it: speak through her, set the lights and the camera, dress her, pose her, start a dance, and capture the moment. Your agent gets a body, a face, and a voice.

Connect in seconds — click the robot button in the chat bar (or type /mcp bind), scan the QR code with your phone, and send one message to your agent. No API keys, no configuration files.

Your agent speaks as her — hand the conversation to your agent: your words go to it, and she answers with its words, lip-synced and with matching expressions. Switch back to her built-in AI anytime.

A director at your command — more than 40 tools: camera angles, studio lighting, the look of the picture, poses, motions, outfits, hairstyles, backdrops, face sculpting, and her voice, all just by asking.

She remembers you — your agent can tell her your name, what you like, and what is going on in your life, and she keeps it in mind in her own conversations too.

Photos, video, and live streams — ask for a picture, record a clip, or go live on YouTube; your agent can look at the screen to check the result.

On your PC or from the cloud — local agents connect right on this PC; cloud agents such as Muse reach your game through our secure relay. The game connects out, so nothing on your PC is opened to the internet.

You stay in control — one command issues a new key and locks out every agent using the old one, and the relay only passes messages through: it does not keep your conversations.

Key Features

Real-time AI conversation — open-ended dialogue by voice or text, answered aloud sentence by sentence

A performance that feels alive — 74 facial expressions, accurate lip-sync, and gestures chosen for each sentence; when you are quiet, she still moves, glances, and breathes

10 voice languages, 31 interface languages — she speaks English, Chinese, Japanese, Korean, German, French, Spanish, Italian, Portuguese, and Russian; the interface follows your system language, with optional dual subtitles in your native language

Voices — nine built-in on-device voices, voices you clone on your own PC, or your own ElevenLabs account

Face sculpting — detailed sliders, control points, and presets, plus skin tone and makeup; or let your agent shape her face from a portrait

Dance, pose, and move — a motion library of dances and idle moves, pose presets, and free posing; she can even learn a new dance from an MMD motion file and its song

Webcam mirror — your webcam drives her face and body in real time

Songs written for you — give her a theme and she writes and sings an original song with karaoke lyrics, or learns a song from your own music file

Studio camera and lighting — orbit freely, frame shots with camera presets, switch lighting moods, take photos, and record video in-game

High-fidelity digital humans — built with Unreal Engine 5: detailed skin, eyes, and simulated hair rendered in real time

You Choose Her Mind

Her intelligence is yours to configure. Run the built-in model directly on your PC — fully offline, completely private, with no subscription and no API keys required. Plug in your own OpenAI, Gemini, or Grok key if you prefer the cloud, or let the AI agent you already trust take over through MCP. Switch whenever you like — and as models and updates improve, so does she: smarter, more natural, more alive.

Make Her Truly Yours

Choose Luna or Orion, change outfits and hairstyles, sculpt every feature of the face, give them the voice you like, and set the backdrop and the light — then watch the personality come through in every glance and gesture. Frame the perfect shot with the free camera and keep it as a photo or a video.

Made for creators, storytellers, AI tinkerers, and anyone curious about what it feels like when AI truly responds.

Steam gallery

Screenshots

10 available
More game data · Seller history