A growing ecosystem of inference-focused startups is drawing strong investor interest, as competitors like Baseten and Fireworks AI have raised hundreds of millions at multibillion valuations.

TechCrunch published the February 11, 2026 report noting that terms could still change and that Modal and General Catalyst did not respond to requests for comment.

CEO Erik Bernhardsson, with background from Spotify and Better.com, leads Modal Labs and drives a focus on resolving GPU-intensive inference bottlenecks to deliver cost savings and faster responses.

Modal was founded in 2021 by Erik Bernardsson and others, with early backers including Lux Capital and Redpoint Ventures, and continues to attract interest from major investors like General Catalyst.

There are talks of a new funding round led by General Catalyst that could value Modal at about $2.5 billion, more than doubling its prior $1.1 billion valuation, with an annualized revenue run rate around $50 million and terms still fluid.

Modal is positioned within this momentum, as part of a broader shift in AI investment from model training toward deployment efficiency and operational performance.

Modal Labs specializes in AI inference infrastructure, aiming to reduce compute costs and latency by optimizing how trained models are run, thereby improving customer experience for AI-powered applications.

Recent moves include vLLM rebranding to Inferact with a $150 million seed round and an $800 million valuation, and SGLang becoming RadixArk with seed financing at a $400 million valuation.

If the round succeeds, it would reinforce a trend toward investing in infrastructure that enables scalable, cost-effective AI deployment, potentially accelerating innovation and competition.

Both Modal and General Catalyst had not commented publicly at the time of reporting.

Market momentum suggests rising demand for compute savings and faster AI responses, which could drive higher valuations for inference-focused companies.

Efficient AI inference is critical for scalable deployment, reducing GPU costs and latency and enabling enterprise-wide AI adoption.