AI inference infrastructure provider Modal Labs is nearing the completion of a $750 million funding round led by Accel. The company’s post-money valuation is expected to be $15.75 billion. According to sources, the new round will more than triple the $4.65 billion valuation Modal reached in the $355 million funding round it announced just four months ago.
The development comes as demand for inference services is increasing, particularly from customers using open-source models. Baseten is reportedly in funding talks at a $26 billion valuation, while Fireworks and Fal, which provides inference services for video and image generation, are also said to be in talks with investors for rounds that would significantly increase their valuations. Although revenues are growing rapidly, companies’ profit margins are low because the costs of purchasing or renting computing capacity are high. Fireworks said in July that its annualized revenue had increased fivefold year over year to $1 billion.
- Modal was founded in 2021 by CEO Erik Bernhardsson and CTO Akshat Bubna.
- The New York-based company has approximately 150 employees.
- The company enables developers to train AI models and run computationally intensive workloads without managing their own servers.
- Its customers include Cognition, Suno, Ramp and Substack.
Modal told Reuters that its annualized revenue had exceeded $300 million as of May. The company said in late July that a customer’s data had been compromised in the same campaign as the attack targeting Hugging Face. Bubna said the incident stemmed not from a vulnerability in Modal’s systems but from a publicly accessible endpoint in the customer’s code, and that the platform had not been compromised.
Background
Labs is not a new name in the FikirPilot archive: we have published 7 news articles mentioning this name in the last 90 days; the latest is dated 29 September 2026.
Term: inference
Inference is when a trained model generates a response to a new input; unlike training, it incurs a new cost each time it is used.