Models

Modal raises 750 million dollars in Series D funding at 15.75 billion dollar valuation

October 17, 2026 · 2 min read

motion blurred blue ocean waves
Sayan Ghosh / Unsplash

Modal Labs, an artificial intelligence infrastructure provider, announced that the company raised 750 million dollars in a Series D funding round. Accel led the round with participation from Kleiner Perkins and other previous investors. The company's valuation now stands at 15.75 billion dollars, which is more than three times the valuation after the previous funding round in May. The revenue growth reflects the rapid expansion of Modal, with the company reporting annual revenue exceeding one billion dollars since the May funding round. Nearly one third of Modal's revenue comes from Sandboxes, a cloud service with containers customized for AI workloads. Developers can use these containers as environments for training models. The service also supports running AI agents and code created by agents. Modal states that Sandboxes can prepare tens of thousands of containers per second. This scalability is made possible by a specifically built infrastructure management system or scheduler. The scheduler processes container preparation requests in parallel across multiple servers to accelerate work.

Modal claims the company can run up to one million containers simultaneously in internal tests conducted recently. Sandboxes that host AI agents sometimes connect to a second set of containers or Sidecars that install supplementary software. For example, a tool that detects and screens for hallucination symptoms may run. Modal states that Sandboxes will run Sidecars on the same physical server as the containers they support to reduce data transmission latency. Some agent types require the ability to access external applications. According to Modal, the Sandboxes feature called Secret Injection allows agents to use application credentials without seeing the data. This access restriction helps protect credentials from unwanted agents and cyber attacks. Modal offers Sandboxes alongside a cloud service customized for AI training loads. This allows developers to activate GPU clusters with relatively simple scripts. Chips in Modal's clusters exchange data using RDMA, a network protocol that accelerates data transfer by reducing the path packages travel.

The company competes in the infrastructure market with a service called Endpoints, which allows developers to deploy AI models on hosted infrastructure and make them available through application infrastructure. If an organization does not want to create its own model, it can choose to use open source algorithm libraries provided by Modal. The company's services run on hardware leased from more than 20 cloud providers. If one cloud provider experiences technical issues, client workloads can continue running on other cloud providers. Modal will use part of the newly raised funds to develop products. The company plans to launch a service focusing on reinforcement learning and post-training processes. This is a step in the AI development workflow designed to improve reasoning ability, safety and cost efficiency of AI being trained.