fal.ai
fal.ai
Provides a unified API to run generative image, video, audio, and 3D AI models and to deploy or train custom models on serverless GPUs and dedicated compute.
The fal.ai API uses API key authentication; has usage based pricing. Verified 2026-08-03 with per-field source links.
Free tier
not documented null
no sourcePricing
GPU hourly: H100 $4.50/h list ($1.89/h as low as), B200 $6.25/h, H200 $4.50/h; Model APIs billed per output, e.g. Wan 2.5 video $0.05/second, Seedream V4 image $0.03/image
fal.ai ↗Rate limits
not documented null
no sourceOpenAPI spec
not documented null
no sourceMCP server
not documented null
no source