AI Industry

Modal Raises $750M at $15.75B as Run Rate Tops $1B

Oct 15, 2026·3 min read·By Rahish Saifi

Photo of copper liquid-cooling pipes and braided hoses meeting at a manifold valve in a data centre, representing Modal's GPU cloud Series D

On October 14, 2026, Modal said it has raised a $750 million Series D at a $15.75 billion post-money valuation. In a post by CEO and co-founder Erik Bernhardsson, Accel led the round; Kleiner Perkins joined, and all of its existing investors participated, including Menlo Ventures, General Catalyst, Redpoint, Lux Capital, Amplify Partners, and Bain Capital Ventures.

The financing lands five months after Modal’s May 21, 2026 Series C. Bernhardsson wrote that usage has tripled since that announcement and Modal has grown more than 15× year-over-year, recently crossing $1 billion in annualized run rate. Accel said Modal crossed $1 billion in revenue run rate during September 2026. The post-money mark of $15.75 billion is roughly 3.4× the $4.65 billion post-money valuation Modal disclosed with its $355 million Series C (Modal).

What Modal builds

Modal sells serverless AI compute: developers write Python, and Modal provisions GPUs, CPUs, storage, and networking behind decorators and APIs. Bernhardsson frames it as a cloud rebuilt for AI—not a single-purpose GPU rental—covering elastic inference, batch jobs, training, and isolated runtimes for untrusted code.

The company rebuilt core infrastructure itself: its own file system, container runtime, scheduler, and networking on a capacity pool spanning 20+ clouds, aiming to avoid YAML-heavy orchestration and long GPU procurement cycles on legacy stacks.

Sandboxes and the agent runtime wave

Modal launched Sandboxes in 2023 for LLM-generated and agent-written code. Bernhardsson said Sandboxes are now a $300 million run-rate product on their own, with close to 3 billion sandboxes spun up, per Modal (Accel puts it at more than 3 billion), and scale to a million concurrent.

Use cases Modal highlights include background agents (Cognition, Ramp, Linear, Hebbia), vibecoding platforms, reinforcement learning, and code execution at companies from DoorDash to Suno. TypeSafe AI’s Jev served 1 trillion tokens on Modal in its first three days, per Bernhardsson.

Funding history

Round	Disclosed size	Valuation	Lead / notes
Series C (May 21, 2026)	$355M	$4.65B post-money	General Catalyst and Redpoint led
Series D (Oct 14, 2026)	$750M	$15.75B post-money	Accel led; Kleiner Perkins joined

In late September, TechCrunch reported, citing a source, that Modal was nearing a $750 million round led by Accel at a $15.75 billion valuation including the investment; Modal declined to comment at the time. The Oct 14 announcement matches those terms.

Why it matters

For platform teams: Modal’s growth is tied to agent sandboxes and burst inference—workloads that do not map cleanly onto reserved GPU clusters or vanilla Kubernetes. If your roadmap includes agents that execute generated code, a managed sandbox tier with million-concurrency claims is directly competitive with rolling your own isolation layer on a hyperscaler.

For finance and strategy leads: Crossing $1B annualized run rate in roughly five months after a $4.65B mark shows how fast inference and agent-runtime spend can re-rate private infrastructure vendors. Accel, which joined Modal's Series C, led the Series D.

For the market map: Modal competes with GPU clouds, hyperscaler ML platforms, and inference specialists for the same scarce GPU capacity. Bernhardsson’s emphasis on owning the scheduler and pooling 20+ clouds is a bet that multi-cloud arbitrage beats single-vendor reservations as agents multiply short-lived jobs. Qualitatively, that pits Modal against both “rent a cluster” vendors and vertically integrated clouds—without requiring customers to standardize on one public cloud.

Team and hiring

Bernhardsson said Modal now has a team of 200+ across New York, San Francisco, and Stockholm. Co-founder Akshat Bubna is named alongside Bernhardsson in Accel’s investment note.

Bottom line

Modal’s $750M Series D at $15.75B post-money, led by Accel, funds a full-stack AI compute platform riding 15×+ growth, $1B+ annualized run rate, and a $300M Sandboxes franchise. The round validates agent runtimes and elastic inference as infrastructure categories—not side features on a general cloud.