Generating Videos

MoneyPrinter supports two methods for generating videos: the web UI and the REST API.

Using the Web UI

1. Access the Frontend

Ensure all services are running:

# Terminal 1: Backend
uv run python Backend/main.py

# Terminal 2: Worker
uv run python Backend/worker.py

# Terminal 3: Frontend
cd Frontend
python3 -m http.server 3000

Open http://localhost:3000 in your browser.

2. Basic Generation

Enter video subject

Type the topic for your video in the Video Subject field:

Top 5 AI Tools for Developers

Select voice

Choose a TikTok TTS voice from the dropdown. Popular options:

Voice Code	Description
`en_us_001`	English (US) - Male
`en_us_006`	English (US) - Female
`en_uk_001`	English (UK) - Male
`es_mx_002`	Spanish (Mexico)
`fr_001`	French

Configure advanced options (optional)

Expand the Advanced Options section:

AI Model: Select from your installed Ollama models
Paragraphs: Number of paragraphs in the script (1-5)
Custom Prompt: Override default script generation
Subtitle Position: center or bottom
Text Color: Hex color code (e.g., #FFFF00)
Background Music: Toggle on/off
YouTube Upload: Enable automatic upload
Threads: CPU threads for rendering (2-16)

Click Generate

Click the Generate button. The UI will:

Submit the job to the API
Receive a jobId
Start polling for status and events
Display real-time progress logs

Monitor progress

The status area shows:

Current job state (queued, running, completed, failed)
Real-time event logs from the pipeline
Error messages (if any)
Download link when complete

3. View Output

When generation completes:

The video is saved as output.mp4 in the project root
A download link appears in the UI
The final video path is shown in the logs

4. Cancel Generation

To cancel a running job:

Click the Cancel button
The worker will stop processing at the next checkpoint
Job state changes to cancelled

Cancellation is not instant. The worker checks for cancellation between pipeline stages.

Using the API

1. Queue a Job

Submit a generation request:

curl -X POST http://localhost:8080/api/generate \
  -H "Content-Type: application/json" \
  -d '{
    "videoSubject": "Top 5 AI coding assistants",
    "aiModel": "llama3.1:8b",
    "voice": "en_us_001",
    "paragraphNumber": 1,
    "customPrompt": "",
    "subtitlesPosition": "center",
    "color": "#FFFF00",
    "useMusic": false,
    "automateYoutubeUpload": false,
    "threads": 4
  }'

Response:

{
  "status": "success",
  "message": "Video generation queued.",
  "jobId": "abc123-def456-ghi789"
}

2. Check Job Status

Query the job state:

curl http://localhost:8080/api/jobs/abc123-def456-ghi789

Response:

{
  "status": "success",
  "job": {
    "id": "abc123-def456-ghi789",
    "state": "running",
    "cancelRequested": false,
    "resultPath": null,
    "errorMessage": null,
    "createdAt": "2026-03-05T19:30:00Z",
    "startedAt": "2026-03-05T19:30:05Z",
    "completedAt": null
  }
}

3. Fetch Job Events

Get real-time progress logs:

curl "http://localhost:8080/api/jobs/abc123-def456-ghi789/events?after=0"

Response:

{
  "status": "success",
  "events": [
    {
      "id": 1,
      "type": "queued",
      "level": "info",
      "message": "Job queued.",
      "payload": null,
      "timestamp": 1709665800.0
    },
    {
      "id": 2,
      "type": "running",
      "level": "info",
      "message": "Job started.",
      "payload": null,
      "timestamp": 1709665805.0
    },
    {
      "id": 3,
      "type": "log",
      "level": "info",
      "message": "[Video to be generated]",
      "payload": null,
      "timestamp": 1709665806.0
    }
  ]
}

Use the after parameter to poll only for new events. Pass the last event ID you received.

4. Poll for Completion

Implement polling logic:

import requests
import time

API_BASE = "http://localhost:8080"
job_id = "abc123-def456-ghi789"
last_event_id = 0

while True:
    # Check job status
    response = requests.get(f"{API_BASE}/api/jobs/{job_id}")
    job = response.json()["job"]
    
    if job["state"] in ["completed", "failed", "cancelled"]:
        print(f"Job {job['state']}")
        if job["resultPath"]:
            print(f"Video saved to: {job['resultPath']}")
        if job["errorMessage"]:
            print(f"Error: {job['errorMessage']}")
        break
    
    # Fetch new events
    events_response = requests.get(
        f"{API_BASE}/api/jobs/{job_id}/events",
        params={"after": last_event_id}
    )
    events = events_response.json()["events"]
    
    for event in events:
        print(f"[{event['level']}] {event['message']}")
        last_event_id = event["id"]
    
    time.sleep(2)  # Poll every 2 seconds

5. Cancel a Job

Request cancellation:

curl -X POST http://localhost:8080/api/jobs/abc123-def456-ghi789/cancel

Response:

{
  "status": "success",
  "message": "Cancellation requested."
}

Custom Prompts

Override the default script generation with a custom prompt:

{
  "videoSubject": "AI in Healthcare",
  "customPrompt": "Write a 30-second script about how AI is revolutionizing medical diagnosis. Focus on early cancer detection. Use simple language suitable for a general audience. Do not include any markdown formatting.",
  "aiModel": "llama3.1:8b",
  "voice": "en_us_001",
  "paragraphNumber": 1
}

Custom prompts completely replace the default system prompt. Ensure your prompt includes formatting instructions.

Advanced Configuration

Subtitle Styling

Customize subtitle appearance:

{
  "subtitlesPosition": "bottom",
  "color": "#00FF00",  // Green text
  "threads": 8  // Faster rendering
}

Subtitle rendering uses the font at fonts/bold_font.ttf:

Backend/video.py

generator = lambda txt: TextClip(
    text=txt,
    font_size=70,
    color=text_color,
    stroke_color="black",
    stroke_width=5,
    font=str(FONTS_DIR / "bold_font.ttf"),
    size=video.size,
    method="caption",
)

Video Dimensions

Videos are rendered at 1080x1920 (vertical format) for YouTube Shorts/TikTok:

Backend/video.py

clip = clip.resized(height=1920)
clip = clip.cropped(width=1080, height=1920, x_center=clip.w / 2, y_center=clip.h / 2)

Background Music

Enable background music:

{
  "useMusic": true
}

Requirements:

Place .mp3 files in the Songs/ directory
Music is mixed at 10% volume
Random song selection

See Background Music Guide for details.

Common Use Cases

Generate a Quick Short

curl -X POST http://localhost:8080/api/generate \
  -H "Content-Type: application/json" \
  -d '{
    "videoSubject": "Fun fact about space",
    "aiModel": "llama3.1:8b",
    "voice": "en_us_001",
    "paragraphNumber": 1
  }'

Educational Content

curl -X POST http://localhost:8080/api/generate \
  -H "Content-Type: application/json" \
  -d '{
    "videoSubject": "How photosynthesis works",
    "aiModel": "llama3.1:8b",
    "voice": "en_us_006",
    "paragraphNumber": 2,
    "subtitlesPosition": "bottom",
    "color": "#FFFFFF"
  }'

Marketing Video

curl -X POST http://localhost:8080/api/generate \
  -H "Content-Type: application/json" \
  -d '{
    "videoSubject": "New product launch",
    "customPrompt": "Create a 20-second promotional script for a new smartwatch with health tracking features. Highlight benefits and call-to-action.",
    "aiModel": "llama3.1:8b",
    "voice": "en_us_001",
    "useMusic": true,
    "paragraphNumber": 1
  }'

Troubleshooting

Job stays in 'queued' state

Ensure the worker is running: uv run python Backend/worker.py
Check worker logs for errors
Verify database connectivity

No stock videos found

Verify PEXELS_API_KEY in .env
Try a broader video subject
Check Pexels API quota (200 requests/hour on free tier)

TTS fails

Verify TIKTOK_SESSION_ID is valid
Session IDs expire - refresh if needed
Check for special characters in script

Subtitle rendering error

Ensure ImageMagick is installed
Set IMAGEMAGICK_BINARY in .env if not auto-detected
Check font file exists at fonts/bold_font.ttf

Next Steps

Ollama Models

Choose and configure AI models

Background Music

Add music to your videos

YouTube Upload

Automate YouTube uploads

Pipeline

Understand the video generation process

Get Started

Setup & Configuration

Core Concepts

Guides

Development

Generating Videos

Generating Videos

Using the Web UI

1. Access the Frontend

2. Basic Generation

3. View Output

4. Cancel Generation

Using the API

1. Queue a Job

2. Check Job Status

3. Fetch Job Events

4. Poll for Completion

5. Cancel a Job

Custom Prompts

Advanced Configuration

Subtitle Styling

Video Dimensions

Background Music

Common Use Cases

Generate a Quick Short

Educational Content

Marketing Video

Troubleshooting

Next Steps

Ollama Models

Background Music

YouTube Upload

Pipeline

Build docs developers (and LLMs) love

Get Started

Setup & Configuration

Core Concepts

Guides

Development

​Generating Videos

​Using the Web UI

​1. Access the Frontend

​2. Basic Generation

​3. View Output

​4. Cancel Generation

​Using the API

​1. Queue a Job

​2. Check Job Status

​3. Fetch Job Events

​4. Poll for Completion

​5. Cancel a Job

​Custom Prompts

​Advanced Configuration

​Subtitle Styling

​Video Dimensions

​Background Music

​Common Use Cases

​Generate a Quick Short

​Educational Content

​Marketing Video

​Troubleshooting

​Next Steps

Ollama Models

Background Music

YouTube Upload

Pipeline

Build docs developers (and LLMs) love

Generating Videos

Using the Web UI

1. Access the Frontend

2. Basic Generation

3. View Output

4. Cancel Generation

Using the API

1. Queue a Job

2. Check Job Status

3. Fetch Job Events

4. Poll for Completion

5. Cancel a Job

Custom Prompts

Advanced Configuration

Subtitle Styling

Video Dimensions

Background Music

Common Use Cases

Generate a Quick Short

Educational Content

Marketing Video

Troubleshooting

Next Steps