video-content-pipeline

Faceless video production for YouTube/TikTok/Shorts: Pixar-style images (free via Pollinations), parallax effects, voiceover (Edge TTS cloud), composer — the full pipeline from script to finished video. DISCLOSURE: scene prompts go to image.pollinations.ai, narration text goes to Microsoft Edge TTS (cloud).

Production of faceless videos — from script to finished MP4, without filming.

What you get

1. Pixar-style images — free via Pollinations.ai (Flux, no API key)
2. 2.5D parallax effect — brings still images to life (PIL/ffmpeg)
3. Voiceover — Edge TTS (multiple languages/voices) or Piper (offline)
4. Composer — image + text + audio → finished MP4 (ffmpeg)
5. Batch production — whole folder → all videos

Quick start

bash

# 1) Script → scenes
python3 scripts/scene_plan.py script.txt scenes.json

# 2) Generate images (free)
python3 scripts/generate_images.py scenes.json images/

# 3) Voiceover
python3 scripts/voiceover.py scenes.json audio/

# 4) Assemble video
python3 scripts/compositor.py scenes.json images/ audio/ final.mp4

Files

text

video-content-pipeline/
├── SKILL.md
└── scripts/
    ├── scene_plan.py      # split script → scenes with prompts
    ├── generate_images.py # Pollinations (Flux, free)
    ├── voiceover.py       # Edge TTS / Piper
    └── compositor.py      # ffmpeg: images + text + audio → MP4

Specs (platform)

• YouTube: 1920×1080 · TikTok/Shorts: 1080×1920
• 30-60 sec per video · 3-5 sec per scene
• Voiceover: 150 words/min

Monetization (niche ideas)

• Islamic content (the FactSage model)
• Stories/facts · DIY · motivational quotes

Generate cinematic scene prompts for your video automatically:

bash

# 2) Generate scenes (PAID call — costs per call)
python3 scripts/scene_gen.py "Islamic history coffee culture"

• ⚠️ Paid call — each run charges your key. The free pipeline above remains free.
• 🔒 PRIVACY: your topic is sent to the external API.

---