J
Jim CaiOct 16, 2026
Generative Media for Developers
8 min readHow to Fine-Tune Wan 2.2 With an API
A step-by-step guide to training a Wan 2.2 LoRA on fal with cURL and Python: dataset layout, the trainer endpoint and its inputs, the published per-step price, and calling the trained LoRA from the Wan 2.2 text-to-video endpoint.
Serverless AI Inference
9 min readNVIDIA B200 Price in 2026: Specs, Hourly Rates and Reserved Capacity
The NVIDIA B200 pairs 192 GB of HBM3e with 8 TB/s of memory bandwidth and native FP4. This guide covers its specs, the three ways B200 time is priced (hourly, per second, reserved), and fal's published B200 rate with worked cost examples.
J
Jim CaiOct 16, 2026