By Hamza Faisal

Modal, the cloud platform that runs production AI workloads for companies like DoorDash, Cognition, and Runway, has raised a $750 million Series D at a $15.75 billion post-money valuation, the company announced on Wednesday. The round was led by Accel, with Kleiner Perkins joining and all of the startup's existing investors participating.

The new valuation is more than triple the $4.65 billion mark Modal set just five months ago, when it raised a $355 million Series C. Behind the leap: Modal says it recently crossed $1 billion in annualized revenue run rate, with usage tripling since May and year-over-year growth of more than 15 times.

Chief executive Erik Bernhardsson said demand has been accelerating since the Series C, and that the company crossed the $1 billion run-rate milestone in September with no sign of slowing down. The company employs more than 200 people across New York, San Francisco, and Stockholm.

Sandboxes carry a third of the revenue

Nearly a third of Modal's revenue now comes from a single product: Sandboxes, a cloud service that spins up hosted containers optimized for AI workloads. Developers use them as training environments and, increasingly, as the place where AI agents run the code they generate.

Modal says it has spun up close to 3 billion sandboxes to date and can scale to a million concurrent containers, using a custom scheduler that parallelizes provisioning requests across servers.

The Sandboxes business alone accounts for roughly $300 million of run-rate revenue. TypeSafe AI's Jev model processed one trillion tokens on Modal in its first three days, and the customer list spans video generation at Runway, robot training at Physical Intelligence, and drug discovery at Chai Discovery.

A full rebuild across 20-plus clouds

Modal says it has rebuilt its core stack — file system, container runtime, scheduler, and networking — to run across more than 20 cloud providers.

Recent releases include multi-node training with RDMA-connected clusters, Modal servers for HTTP serving, and new features for large language model inference, with virtual-machine-based Sandboxes now generally available. The company says it is moving further up the stack into reinforcement learning and post-training products.

Accel said the infrastructure to train, serve, and run AI is the fastest-growing market in technology, and that the industry is only in the very early stages of building it. Existing backers Menlo Ventures, General Catalyst, Redpoint, Lux Capital, Amplify Partners, and Bain Capital Ventures all participated in the new round.

The infrastructure bet behind the agents boom

The financing is a bet that the real control point in AI is not the models but the layer where they actually run. As AI agents move from demos to production workloads — logging into apps, writing code, filtering their own hallucinations — someone has to provision millions of isolated, secure containers every second.

Modal's pitch is that it already does.

Whether that justifies a $15.75 billion price tag depends on whether agent-driven demand keeps compounding, or whether the current GPU-scarcity moment is doing the heavy lifting. For now, investors are paying for the curve: 15-times year-over-year growth, and a revenue run rate that tripled in five months.

By Hamza Faisal