fal

About

Hosted inference platform with 1,000+ image, video, audio and 3D models behind one API key. Most models are billed per output from prepaid credits. You can also deploy your own models on serverless GPUs.

What you can do with it

• Generate images with FLUX and other hosted models, try-on and photo restoration too
• Turn images into videos with Kling or Sora 2, or lip sync a face to new audio
• Deploy your own model on GPUs that autoscale from zero

Get started

1. Create an API key at fal.ai/dashboard/keys
2. Export it as FAL_KEY
3. Buy prepaid credits

Example

curl -X POST "https://queue.fal.run/fal-ai/flux/schnell" \
  -H "Authorization: Key $FAL_KEY" \
  -H "Content-Type: application/json" \
  -d '{"prompt": "a futuristic cityscape at sunset"}'

Details

Hosting

Hosted service

Available in

Worldwide

Official SDKs

Python, JavaScript/TypeScript, Swift, Kotlin, Java, Dart

MCP server

Remote