Use Case
Generate Talking Head & UGC Videos with AI — From Your Terminal
Transform any script into a complete UGC video in one command. The pipeline handles scene splitting, AI talking heads, AI B-roll generation, TTS voiceover, crossfade assembly, animated subtitles, background music, and end-screen CTA automatically.
Best Models for This
These models are recommended for generating talking head & ugc videos. Each one brings different strengths to this workflow.
| Model | Why | Credits | CLI Command |
|---|---|---|---|
| ugc-basic | Full pipeline (AI talking heads + AI B-roll + AI voiceover). Handles scene splitting, crossfade assembly, animated subtitles, background music, and end-screen CTA automatically. | 500–1,500 per 5–15s video | agent-media ugc "Ever wonder why top founders wake up at 5am?" --sync |
Step-by-Step Workflow
Follow these steps to start generating talking head & ugc videos with agent-media CLI.
Generate from script
Provide your script and the UGC pipeline splits it into scenes, generates TTS voiceover, creates AI talking head clips and AI B-roll, then assembles everything with crossfades, animated subtitles, background music, and an end-screen CTA.
$ agent-media ugc "Ever wonder why top founders wake up at 5am?" --sync
Add a face photo for talking heads
Provide a face photo to drive talking head scenes. The pipeline auto-detects the best TTS voice to match the face and uses AI lip-sync for realistic results.
$ agent-media ugc "script..." --face-url ./photo.png --sync
Customize style and music
Choose from subtitle styles (hormozi, minimal, bold, karaoke, clean), set the background music genre, and customize aspect ratio and duration to match your brand.
$ agent-media ugc "script..." --style hormozi --music chill --sync
Multi-scene with custom B-roll images
Supply your own images as B-roll sources. The pipeline animates them using AI video generation and weaves them into the assembled video between talking head scenes.
$ agent-media ugc "script..." --broll-images "url1,url2" --sync
Dub into another language
Generate the full UGC video in a different language. The pipeline translates the script, synthesises a matching TTS voice, and re-assembles the video with localized subtitles.
$ agent-media ugc "script..." --dub-language es --sync
Create a reusable persona
Save a persona with a specific voice and face for consistent UGC across videos. Reference it with --persona brand in future ugc commands to maintain a unified creator identity.
$ agent-media persona create --name brand --voice ./sample.mp3 --face ./photo.png
How Much Does It Cost
A 5-second UGC video costs 150 credits, covering AI talking heads, voiceover, animated subtitles, and full assembly. A 10s video is 300 credits, 15s is 450 credits. The Creator plan ($39/month for 3,900 credits) lets you produce ~13 videos at 10s per month. For higher volume, Pro Plus ($129/month for 12,900 credits) covers ~43 videos monthly.
Related Use Cases
TikTok & Reels Content
Short-form vertical video dominates social media. TikTok, Instagram Reels, and YouTube Shorts all require 9:16 portrait content that grabs attention in the first second. With agent-media CLI, you can generate scroll-stopping short-form videos directly from your terminal — no video editor, no stock footage library, no browser tabs. Pick a model, write a prompt, and publish.
YouTube Shorts Automation Pipeline
Publishing YouTube Shorts consistently is a volume game — the algorithm rewards frequent uploads. Manually creating each Short is a bottleneck. agent-media CLI is a terminal tool, which means it integrates directly into shell scripts, cron jobs, and CI/CD pipelines. You can build an automated pipeline that generates, reviews, and queues Shorts without manual intervention for every video.
Product Demo Videos
Product demo videos convert browsers into buyers, but hiring a video production team for every SKU is expensive and slow. AI video generation lets you create animated product showcases, explainer clips, and walkthrough videos from a text description. With agent-media CLI, you describe your product scene, choose a model optimized for object detail and smooth motion, and get a polished clip back in minutes.