FFMPEG API Guide

How do I generate AI video captions with an API?

Generate TikTok-style word-by-word captions from video URLs with transcription and Remotion rendering.

Get API Key API Docs MCP Server OpenAPI JSON

AI Captions

Extracts audio, transcribes it with word timestamps, and renders styled captions into a video. Also returns caption artifact URLs when available.

Endpoint: POST /api/videos/add-tiktok-captions

Authentication: send X-API-Key: your_api_key_here.

curl Example

curl -X POST "https://ffmpegapi.net/api/videos/add-tiktok-captions" \
  -H "X-API-Key: your_api_key_here" \
  -H "Content-Type: application/json" \
  -d '{
  "video_url": "https://example.com/video.mp4",
  "subtitle_style": "yellow-bg",
  "position": "bottom"
}'

Request Body

{
  "position": "bottom",
  "subtitle_style": "yellow-bg",
  "video_url": "https://example.com/video.mp4"
}

Success Response

{
  "captions_json_url": "https://ffmpegapi.net/api/storage/auto-captions/captions_abc123.json",
  "download_url": "https://ffmpegapi.net/api/storage/auto-captions/video_abc123.mp4",
  "message": "Video with auto-generated TikTok captions rendered successfully",
  "srt_url": "https://ffmpegapi.net/api/storage/auto-captions/captions_abc123.srt",
  "success": true,
  "vtt_url": "https://ffmpegapi.net/api/storage/auto-captions/captions_abc123.vtt",
  "word_count": 142
}

Use From AI Agents

Agents can call this workflow from REST using the OpenAPI spec or from the hosted MCP server at https://mcp.ffmpegapi.net/mcp. Keep API keys in request headers, not tool arguments.

FAQ

FFMPEG API exposes POST /api/videos/add-tiktok-captions to transcribe video audio and render animated captioned video outputs.

Use async: true for larger media, caption rendering, or jobs likely to exceed a normal HTTP timeout. Poll job status until completion.