AI Image Studio
Better prompt controls and a current, authenticated image-engine connection. Your API key stays in this page's memory only.
Recent images
Amharic & English Voice Studio
Generate real MP3 narration with neural voices, listen, and download it. Amharic voices are not taken from the phone's installed voice list.
Long-form Video Studio
Turn an Amharic or English script into timed caption scenes, add visuals and narration, then render a longer video. No preset 6/10/15-second cap.
Text-to-video storyboard (Amharic + English)
Talking Avatar Studio
Create talking presenters from portrait photos or uploaded videos, synchronized with Amharic or English narration.
1. Choose your presenter
Your selected photo or video preview will appear here.
2. Add narration
3. Lip-sync engine
Settings
Personalize your Mariham Creator Studio experience.
Mariham Creator Studio
Create beyond imagination.
About the application
Mariham Creator Studio is an AI-powered creative workspace for producing images, natural voice narration, and videos from your mobile device.
Generate creative images using Cloudflare FLUX.
Create Amharic and English neural voice narration.
Build narrated videos with images, captions, and scene transitions.
1.0.0 — International Edition
Yohannes AI
© 2026 Mariham Creator Studio. All rights reserved.
✨ CineFlow AI
Create videos from a text prompt or a reference image. LTX providers generate AI motion; Mariham Motion creates a locally rendered camera-movement video from one image.
Mariham Motion uses your existing FLUX image engine for text prompts and FFmpeg on this device for camera movement. It does not animate subjects independently. Add narration or music below to include audio.
🎙️ Voice narration / music (optional)
Create Amharic or English speech here, use the latest Voice Studio MP3, or upload your own audio. Mariham adds it to the finished video.
👄 AI Lip Sync — Talking portrait
Uses a real AI lip-sync model to animate the visible face to your narration. The output length follows the audio. Each request is limited to 5–15 seconds to control cost.
Choose a 5–15 second audio track to calculate the estimated price. At 480p, 5 seconds costs about $0.25.
Cost estimates use the published per-second rates and are approximate; check your fal.ai billing before generating. Maximum estimated model cost for a 15-second clip at 2K is $4.80.
Privacy: the portrait and audio are uploaded to fal.ai for processing. fal CDN files are publicly readable by anyone with their URL by default; this integration requests a one-hour expiry. Only use faces and voices you own or have permission to animate.
No request is sent unless the server has your FAL_KEY configured, you check the cost-confirmation box, and you press Generate. The API key stays in Termux and is never sent to the browser.
Mariham Motion uses the existing Cloudflare FLUX image configuration only for Text to Video; rendering and camera movement run locally through FFmpeg without a video GPU or new video-provider account. Image to Video animates the uploaded still image with camera movement. This is not frame-by-frame AI animation. You can generate Amharic/English narration or attach audio below; LTX providers also accept an optional audio track.
