Ollama raised $65 million in a Series B round in early July 2026, led by Theory Ventures, to expand its platform for running AI models directly on local machines rather than through cloud APIs.
The company has built one of the most widely adopted tools among developers who want to run open-source language models on their own hardware, whether for privacy, cost control, offline use, or simply more direct control than a hosted API allows. That developer-first adoption has made Ollama a default choice in a specific but fast-growing niche: local-first AI infrastructure, as opposed to the cloud-hosted model layer that dominates most enterprise AI spending.
The round reflects a smaller but persistent theme running alongside 2026's massive cloud AI infrastructure deals -- not every workload needs, or wants, to run through a hyperscaler's API. Developers, researchers, and increasingly some enterprises value the ability to run models locally for cost, latency, and data control reasons, and Ollama has positioned itself as the easiest way to do that.
For founders building developer tools, Ollama's raise is a reminder that strong grassroots adoption among developers -- even without enterprise sales motion -- can be its own durable path to a fundable, category-defining business.