fal
fal is a generative media platform that provides developers with a proprietary inference engine and serverless GPU infrastructure to deploy and scale AI models for image, video, audio, and 3D generation. The platform offers access to over 1,000 pre-optimized models through a unified API, with performance up to 4x faster than alternatives and cost reductions of up to 10x. Trusted by over 2.5 million developers, fal powers production workloads for companies including Canva, Quora, and Perplexity.
- Artificial Intelligence
- Developer Tools
- Software Only
- Creator / Creative Economy
Funding
Raised to date
$2BRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
across 7 rounds
+22$1.3B · Series F
Announced Nov 2026 · from article date
Disclosed cumulative: $2B
Founders
Founder details are not available yet.
Product
Problem
Developers face significant challenges in deploying and scaling generative media models due to the high computational costs and infrastructure complexities associated with diffusion models. Existing solutions often lack the speed, cost-effectiveness, and scalability required for real-time applications and high-volume production workloads.
Solution
fal provides a generative media platform that enables developers to customize, deploy, and scale generative AI models with optimized performance. The platform leverages a proprietary inference engine specifically designed for diffusion models, delivering significant speed improvements and reducing costs and latencies by up to 10x. fal's serverless GPU infrastructure allows users to run models and only pay for the computing resources consumed, optimizing cost efficiency. By abstracting away infrastructure management, fal allows developers to focus on building and deploying creative applications. The platform offers access to over 1,000 optimized models through a unified API, or developers can deploy their own custom models on the same infrastructure, with support for private model endpoints and enterprise-grade compliance.
Target Audience
fal's primary customers are developers and organizations building generative AI applications, including entertainment studios, advertising agencies, e-commerce brands, game developers, and marketing teams. The platform serves both individual developers and large enterprises requiring production-scale AI infrastructure, with customers including Canva, Quora, and Perplexity.
Features
- Optimized inference engine for diffusion models, achieving up to 4x faster performance compared to alternative solutions and reducing costs and latencies by up to 10x
- Serverless GPU infrastructure that scales on demand, eliminating the need for manual resource provisioning, with GPUs including RTX PRO 6000, H100, H200, B200, and B300
- Access to over 1,000 pre-optimized models through a unified API, including image, video, audio, and 3D generation models from labs like Black Forest Labs, Google, OpenAI, and xAI
- Support for fine-tuning models with a LoRA trainer, enabling personalization and style customization in minutes
- Client libraries for JavaScript, Python, and Swift, facilitating integration into existing applications
- Billing based on actual compute usage or model output, providing cost transparency and control, with GPU pricing starting at $1.99 per hour
- SOC 2 compliant with enterprise-grade features including private model endpoints, reserved capacity options, and forward-deployed Applied Machine Learning Engineers
- Platform handles over 100 million inference requests daily with 99.99% uptime for mission-critical reliability