fal Raises $1.65B Across Series E, F

This round signals that generative-media infrastructure is becoming a strategic platform race, not merely an API business.

• –fal spans model access, deployment, inference, and compute, giving developers a single path from prototype to production.
• –NVIDIA’s continued backing highlights the importance of GPU supply, serving efficiency, and inference economics.
• –Developers gain faster access to new image, video, audio, and 3D models through one API, but may also face greater provider lock-in.
• –The next competitive battleground will be reliability, pricing, observability, and latency—not just model count.
• –fal says its infrastructure supports more than 1,500 models and 100M+ daily inference calls.