Image
Stills, edits, upscale, restore
- Qwen-ImageText to image on weights you host.
- Qwen-Image-EditEdit an existing still with a prompt.
- Real-ESRGANUpscale a still.
- CodeFormerRestore faces in a still.
- BiRefNetRemove the background.
Now
What is on.
A plain list of systems you can use on this site today. Generation rooms are open; finished production media waits on connected engines.
Development Mode · labeled preview until engines are connected
Live means the system itself is real. Preview means you can run the room and the listed engine, and the file that returns is a labeled stand-in until GPU and weights are attached.
You can use these now
- Sign inEmail, Google, and X. Real accounts.Live
- StudioOne composer for every creation room.Live
- LibraryJobs and outputs stay in your workspace.Live
- API keysHashed keys. Same jobs as HTTP.Live
- Usage ledgerCredits for members. Unlimited for the house.Live
- House portalEdit pages, images, and video. Open every door.Live
- House metricsJobs, latency, credits, engines — counted from this studio.Live
- Prompt craftAdvanced briefs by medium.Live
- Model libraryEvery listed engine, with license and role.Live
Creation rooms — on, preview output
- Image CreationTurn a thought into a finished still — portraits, fashion, products, worlds. Engines: Qwen-Image, Qwen-Image-Edit, Real-ESRGAN, CodeFormer, BiRefNet.Preview
- Video CreationGive stills and prompts cinematic motion. Engines: LTX-Video, Wan 2.2, HunyuanVideo 1.5, CogVideoX-5B.Preview
- Lip-Sync & PerformanceMake a character speak with natural mouth motion. Engines: MuseTalk, LatentSync 1.6, Hallo2, LivePortrait, InfiniteTalk.Preview
- VoiceWrite a line. Hear it spoken in a voice you choose. Engines: Qwen3-TTS, CosyVoice 3, Chatterbox, Kokoro, RVC.Preview
- MusicOriginal songs, beds, and themes from a short brief. Engines: ACE-Step 1.5.Preview
- Sound DesignEffects, textures, and stems that finish the world. Engines: Stable Audio Open, Demucs.Preview
- CaptionsTurn speech into clean, editable text. Engines: faster-whisper.Preview
Engine roster
All of these are selectable in the studio today. Output is a labeled preview until weights and GPU are connected.
Video
Short clips from text or a still
- LTX-VideoFast short clips from text or a still.
- Wan 2.2Open video quality leader, Apache-2.0.
- HunyuanVideo 1.5Cinematic motion, FP8 on a 4090.
- CogVideoX-5BPrompt-faithful short clips.
Lip-sync
Match a mouth to audio you own
- MuseTalkReplace the mouth on footage to match audio.
- LatentSync 1.6Sharper 512px mouth replace.
- Hallo2Still photo plus audio becomes a talking person.
- LivePortraitAnimate a still from a driving video.
- InfiniteTalkLonger talking video on Wan weights.
Voice
Speech from a script, or convert a voice
- Qwen3-TTSText to speech with a 3-second clone.
- CosyVoice 3Streaming TTS and clone.
- ChatterboxFast MIT-licensed speech.
- KokoroTiny English TTS.
- RVCConvert one voice into another.
Music
Songs and beds on ACE-Step
- ACE-Step 1.5Full songs with vocals from text.
Sound
SFX, textures, stems
- Stable Audio OpenShort SFX and textures. Read the revenue cap.
- DemucsSplit vocals, drums, bass, other.
Captions
Transcribe audio you upload
- faster-whisperSpeech to text for captions and dubbing.