Demo Video

▶ Watch Demo

Prism Motion serves all kinds of video needs for pharma companies — from explainer videos and drug mechanism animations to sales enablement and compliance content. It leverages Manim for programmatic scientific animations, Remotion for React-based video composition, Pexels + Pixabay APIs for stock footage, and SadTalker AI for photorealistic talking face generation. No editors, no agencies, no weeks of back-and-forth — just describe what you need and get a production-ready video.

# Clone the repository
git clone https://github.com/Goyam02/Prism-Motion.git
cd prism-motion

# Backend setup
cd backend
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate
pip install -r requirements.txt

# Frontend setup
cd ../frontend
npm install
npm run dev

# Remotion setup
cd ../remotion
npm install

• Python 3.10+
• Node.js 18+
• FFmpeg
• OpenAI API key
• Pexels API key
• SadTalker (for avatar generation)

Landing Page Hero

Landing Page — Why Prism Motion

Signup — Step 1

Signup — Step 2

Signup — Step 3

Signup — Step 4

Signup — Step 5

Agent Mode — Landing

Agent Mode — Video Generation

Agent Mode — Video Generated

Creator Mode — Landing

Creator Mode — Edit the Pipeline

Creator Mode — Video Generated

POST /api/v1/agent/video/generate

{
  "prompt": "Explain how Metformin works for Type 2 diabetes patients",
  "target_audience": "patients",
  "video_length": "medium",
  "avatar_image": "https://example.com/doctor-portrait.jpg"
}

{
  "job_id": "vid_abc123",
  "estimated_time": "4 minutes",
  "status": "queued"
}

GET /api/v1/agent/video/status/{job_id}

{
  "job_id": "vid_abc123",
  "status": "processing",
  "progress": 67,
  "scenes": [
    { "name": "Introduction", "status": "complete" },
    { "name": "Mechanism of Action", "status": "rendering" },
    { "name": "Benefits", "status": "pending" }
  ]
}

GET /api/v1/agent/video/download/{job_id}

Returns the final rendered MP4 file.

POST /api/v1/creator/pipeline/create

{
  "prompt": "Create a promotional video for our new hypertension medication"
}

{
  "pipeline_id": "pipe_xyz789",
  "pipeline": {
    "scenes": [
      {
        "id": 1,
        "type": "stock",
        "pexels_query": "doctor checkup",
        "duration": 5
      },
      {
        "id": 2,
        "type": "animation",
        "manim_scene": "blood_flow",
        "duration": 8
      }
    ],
    "narration": ["Welcome to our new treatment..."],
    "suggestions": {
      "footage": ["medical office", "patient consultation"],
      "animations": ["cell mechanism", "drug absorption"]
    }
  }
}

PUT /api/v1/creator/pipeline/{pipeline_id}

{
  "scenes": [
    {
      "id": 1,
      "type": "stock",
      "pexels_query": "hospital hallway",
      "duration": 5
    }
  ],
  "narration": ["Updated narration text here..."]
}

{
  "pipeline_id": "pipe_xyz789",
  "status": "updated"
}

POST /api/v1/creator/pipeline/{pipeline_id}/render

{
  "job_id": "vid_abc123",
  "status": "rendering"
}

POST /api/v1/avatar/generate

{
  "portrait_image": "https://example.com/doctor.jpg",
  "audio_file": "https://example.com/narration.mp3"
}

{
  "job_id": "avatar_123",
  "video_url": "https://cdn.prismmotion.ai/avatars/avatar_123.mp4"
}

GET /api/v1/video/status/{job_id}

{
  "job_id": "vid_abc123",
  "status": "complete",
  "progress": 100,
  "video_url": "https://cdn.prismmotion.ai/videos/vid_abc123.mp4"
}

WebSocket /ws/{session_id}

Real-time progress updates:

{
  "type": "progress",
  "status": "processing",
  "scene": "Mechanism of Action",
  "percent": 45
}

Agent Mode — Fully automated video generation. Just describe what you need and let AI handle everything from script to final render. Perfect for quick turnarounds and high-volume production.

Creator Mode — Full creative control with AI assistance. Generate a pipeline, then tweak scenes, swap footage, adjust narration, and preview before rendering. Ideal for precision work and brand compliance.

• Manim — Programmatic scientific animations for MoA videos
• Remotion — React-based video composition and rendering
• Pexels + Pixabay APIs — Stock photo and video footage
• SadTalker AI — Photorealistic talking face generation
• ElevenLabs / OpenAI TTS — Natural voice narration
• FastAPI — High-performance Python backend
• React + TypeScript — Modern frontend UI
• WebSockets — Real-time progress updates

Create a .env file in the backend directory:

# OpenAI Configuration
OPENAI_API_KEY=sk-...

# Pexels API
PEXELS_API_KEY=...

# Pixabay API (optional)
PIXABAY_API_KEY=...

# SadTalker URL (if using remote instance)
SADTALKER_URL=http://localhost:8002

# Database
DATABASE_URL=sqlite:///./prism_motion.db

# Frontend URL (for CORS)
FRONTEND_URL=http://localhost:5173

# Terminal 1: Start backend
cd backend
source venv/bin/activate
uvicorn app.main:app --reload

# Terminal 2: Start frontend
cd frontend
npm run dev

# Terminal 3: Start Remotion (for rendering)
cd remotion
npm run dev

The backend runs on http://localhost:8000, frontend on http://localhost:5173.

# Build frontend
cd frontend
npm run build

# Backend uses uvicorn with workers
cd backend
uvicorn app.main:app --workers 4

prism-motion/
├── backend/
│   ├── app/
│   │   ├── main.py              # FastAPI application
│   │   ├── creator_mode.py      # WebSocket creator mode
│   │   ├── stages/             # Video generation pipeline
│   │   │   ├── stage1_scenes.py
│   │   │   ├── stage2_remotion.py
│   │   │   ├── stage3_script.py
│   │   │   ├── stage4_tts.py
│   │   │   └── stage5_render.py
│   │   ├── moa_stages/         # Manim MoA pipeline
│   │   ├── doctor_ad_stages/   # Doctor advertisement pipeline
│   │   ├── social_media/       # Social media pipeline
│   │   └── utils/              # LLM, Pexels, helpers
│   └── requirements.txt
├── frontend/
│   ├── src/
│   │   ├── components/         # React components
│   │   ├── App.tsx
│   │   └── index.tsx
│   ├── package.json
│   └── images/                 # UI screenshots
├── remotion/
│   ├── src/
│   │   ├── PharmaVideo.tsx    # Main video component
│   │   ├── ComplianceVideo.tsx
│   │   └── Root.tsx
│   └── package.json
└── README.md

Connect to receive live progress:

const ws = new WebSocket('ws://localhost:8000/ws/creator');

ws.onmessage = (event) => {
  const data = JSON.parse(event.data);
  console.log(`${data.scene}: ${data.percent}%`);
};

Query supported regions for TTS and content:

GET /supported-regions

Video rendering fails

• Check FFmpeg is installed: ffmpeg -version
• Verify sufficient disk space for temp files
• Check Remotion/Manim logs in backend output

TTS not working

• Verify OpenAI API key has credits
• Check ElevenLabs API key if using that provider

Stock footage not found

• Try different search queries
• Pixabay fallback will attempt alternative searches

Avatar generation slow

• SadTalker requires GPU for fast processing
• Consider using remote SadTalker instance

Proprietary — All rights reserved.