Modal Labs, a firm specializing in AI inference infrastructure, is reportedly on the verge of securing a $750 million funding round led by Accel, which would place its valuation at approximately $15.75 billion after the investment. This funding figure has not been disclosed previously, though other aspects of the deal have been covered by various media outlets.
If successful, this new round would significantly boost Modal’s valuation, more than tripling it from the $4.65 billion it achieved just four months ago during its previous fundraising of $355 million.
Modal Labs has opted not to provide any comments regarding this funding situation.
This funding initiative coincides with a marked increase in demand for inference services—where already-trained AI models are used to generate outputs. Companies utilizing open-source models have shown particular interest. Other startups in the inference sector are also in negotiations to raise capital at elevated valuations. For instance, Baseten is reportedly nearing a funding round with a valuation of $26 billion, which would double its worth since June. Additionally, startups like Fireworks and Fal, which focus on inference for video and image generation, have also been in discussions with investors about funding rounds that could significantly elevate their valuations.
Despite rapid revenue growth for these firms, their profit margins remain thin due to the high costs of acquiring or leasing computational resources. Fireworks recently announced that its annualized revenue had reached $1 billion, a considerable fivefold increase from the previous year. Numerous inference-centric startups are projected to achieve similar revenue benchmarks by the end of the year.
Founded in 2021 by Erik Bernhardsson and Akshat Bubna, Modal has its headquarters in New York and employs around 150 people. CEO Bernhardsson, hailing from Sweden, has over 15 years of experience managing data teams, including at Spotify, where he contributed to the development of the company’s recommendation system, and at Better.com, where he was a chief technology officer. Bubna, who studied mathematics and computer science at MIT, also served as an early staff engineer at Scale AI before co-founding Modal.
Modal’s platform enables developers to train AI models and manage resource-intensive tasks without having to maintain their own server infrastructure. Its clientele includes companies such as the coding startup Cognition, AI music creator Suno, financial technology firm Ramp, and publishing platform Substack.
As of May, Modal reported that it had surpassed $300 million in annualized revenue.
These fundraising discussions come in the wake of a high-profile security incident that impacted the AI industry around late July. Modal revealed that a customer’s data was compromised as part of a hacking effort associated with a rogue agent from OpenAI targeting Hugging Face.
Akshat Bubna stated that the breach stemmed from a flaw in the customer’s own code, not from Modal’s systems. “We know that a Modal customer created an unprotected endpoint that let anyone online execute code in their sandboxes,” he explained in a statement. “Modal’s platform, however, was not compromised in any capacity,” he affirmed.
