Ollama raises $88M as open models go mainstream

Ollama says it has raised $88 million as demand for open models accelerates. In a post by co-founders Jeff and Michael on July 9, 2026, the company said its platform now serves 8.9 million developers and is used by 85% of the Fortune 500.

The founders framed Ollama as a way to make open models as easy to use as proprietary ones exposed through an API. Their pitch is familiar to anyone who followed their earlier work: the pair previously started Kitematic, which was acquired by Docker in 2015. That work became Docker Desktop, launched in 2016 and now used by more than ten million developers worldwide, according to the post.

Ollama started with a simple goal: let developers download an app, run the latest open models with a single command, and build on top of them through a simple API. The company says that means no permission requests, no API key, and no expensive server hardware to get started.

It says the product is built around three ideas:

• Ownership: developers can keep, customize, and optimize the models they depend on
• Affordability: running models on local hardware avoids per-token billing
• Privacy: data can stay on the local machine, with the same trust model extending to cloud use

The company also used the announcement to argue that open models have moved beyond experimentation. Ollama says its cloud offers access to models including GLM, Nemotron, DeepSeek, Kimi, MiniMax and others, and that token volume there has, on average, more than doubled every month.

Recommended reading

Some AI problems stay unsolvable, even with infinite data

The round includes Peter Fenton at Benchmark, Tomasz Tunguz at Theory Ventures, and Alex Kolicich at 8VC, alongside angel investors including Solomon Hykes, Aaron Katz, Spencer Kimball, Quinn Slack, Marianna Tessel, Michael Montano, plus Y Combinator, Garage Capital, Pace Capital, 49 Palms, and GTMFund.

Ollama said the new funding will go toward seamless hybrid inference, day-one support for newly released open models, and expanding a cloud service that gives teams access to powerful models without giving up ownership or privacy.