Keyboard shortcuts

Press ← or → to navigate between chapters

Press S or / to search in the book

Press ? to show this help

Press Esc to hide this help

Fal.ai

Serverless AI inference platform / API gateway for running generative models – image generation, video generation, upscaling, and other model workloads – on demand, often as an alternative to running models locally.

On One Small Crawl, Fal was referenced once, on 2026-06-08-frames-call, alongside the Gemini shot list (gemini) and the Topaz enhancement path (topaz), in the shot-list / shot-blueprint discussion. Fal is the kind of inference/API layer that can host a LoRA-loaded diffusion model or an upscaling model on demand; whether it was adopted, tested, or merely considered is per-call.

On One Small Crawl

  • 2026-06-08-frames-call – Fal referenced on the call (2026-06-08-frames-call): the shot-list / shot-blueprint discussion mentions Fal alongside Gemini and Topaz. This is consistent with the production evaluating API-based inference (Fal) vs local inference (the local-laptop image-generation agent Cedric ran for a demo on 2026-06-22-frames-call) vs model-in-editor workflows.
  • gemini – Gemini shot-list / shot-blueprint tool
  • topaz – Topaz upscaling / enhancement (local tool path)
  • lora – LoRA character-consistency adapters that could run locally or via an inference API
  • ai-agent-pipeline – ARC + generation pipeline

fal.ai - generative media API (no Wikipedia article; official site)