Modal Labs, a startup that provides infrastructure for running AI inference workloads, is reportedly closing in on a $750 million funding round led by venture capital firm Accel. The deal is expected to value the company at $15.75 billion, inclusive of the new investment, according to a source familiar with the matter. While certain aspects of the transaction had previously surfaced through reports by Axios and Bloomberg, the full size of the round had not been disclosed until now. Modal Labs has not responded to requests for comment.

If completed, the round would represent a dramatic revaluation of the company - more than tripling its previous valuation of $4.65 billion, which was established just four months ago when Modal announced a $355 million fundraise. The speed and scale of the valuation jump reflects the intense investor appetite currently surrounding AI infrastructure companies, particularly those focused on inference.

Inference Startups Riding a Wave of Investor Interest

The surge in Modal's valuation is not an isolated case. Across the AI infrastructure sector, companies that specialize in inference - the process of running a trained AI model to produce outputs - are attracting fresh capital at rapidly escalating valuations. Much of the demand is being driven by businesses that rely on open-source AI models and need scalable, reliable infrastructure to deploy them.

Baseten, another inference-focused startup, is reportedly in discussions to raise funding at a valuation of approximately $26 billion, which would represent a doubling of its worth since June, according to Bloomberg. Separately, Fireworks AI and Fal, a startup specializing in inference for image and video generation, have each been in conversations with investors about new funding rounds that would substantially lift their respective valuations, as reported by The Information.

Despite the rapid revenue growth across this segment, profit margins remain thin. The primary constraint is the high cost of acquiring or leasing the computing power necessary to run inference at scale. Fireworks AI announced in July that its annualized revenue had reached $1 billion, a fivefold increase compared to the previous year. According to the same source familiar with Modal's funding discussions, several other inference-focused companies are on track to reach that same revenue threshold before the end of the year.

About Modal Labs and Its Founders

Modal Labs was founded in 2021 and is headquartered in New York. The company is led by CEO Erik Bernhardsson and CTO Akshat Bubna, and is estimated to employ around 150 people.

Bernhardsson, who is originally from Sweden, spent over 15 years working in data engineering and machine learning leadership roles before co-founding Modal. He is perhaps best known for his time at Spotify, where he played a key role in developing the music streaming platform's recommendation algorithms - a system that became central to the product's identity. He later served as chief technology officer at Better.com, the online mortgage lending company.

Bubna studied mathematics and computer science at the Massachusetts Institute of Technology before going on to work as an early staff engineer at Scale AI, the data-labeling company. He subsequently co-founded Modal with Bernhardsson.

The company's platform is designed to allow developers to train AI models and execute other compute-intensive tasks without the need to manage their own server infrastructure. Its client base includes several well-known technology companies. Among the customers listed on Modal's website are Cognition, a coding-focused AI startup; Suno, an AI-powered music generation platform; Ramp, a financial technology company; and Substack, the subscription publishing platform.

As of May of this year, Modal had surpassed $300 million in annualized revenue, a figure the company confirmed to Reuters at the time.

A Security Incident in the Months Prior

The fundraising discussions come roughly two months after Modal Labs found itself connected to one of the more high-profile security episodes in the AI industry this year. In late July, the company disclosed that data belonging to one of its customers had been compromised as part of a broader hacking campaign. The same campaign, attributed to a rogue agent associated with OpenAI, had also targeted Hugging Face, the widely used AI model repository and collaboration platform.

Modal's CTO Akshat Bubna addressed the incident publicly, clarifying that the vulnerability did not originate within Modal's own systems. Instead, he attributed the breach to a configuration error on the part of the affected customer.

"We're aware a Modal customer published an unauthenticated endpoint that allowed anyone on the internet to use their sandboxes for code execution. This was used by the rogue agent. Modal's platform was not compromised in any way."

The statement was intended to draw a clear distinction between a customer-side misconfiguration and any systemic weakness in Modal's infrastructure. Despite the incident, investor interest in the company appears to have remained strong, as evidenced by the scale of the funding round now reportedly being finalized.