Ollama's $88 million bet that open-weight models will do what Docker did for containers

Ollama announced the round on its blog July 9, framing it around ownership, affordability, and privacy. The company says 8.9 million developers use the platform and that 85% of Fortune 500 companies have some Ollama deployment. Token volume through its cloud offering has more than doubled every month on average. That last stat is the one that matters, growth that steep usually catches investor attention before the curve flattens.

The round drew Benchmark, Theory Ventures, 8VC, and Y Combinator. Docker founder Solomon Hykes and ClickHouse CEO Aaron Katz are advisors. Ollama's pitch maps directly onto Docker's history: Hykes made containers something any developer could run without special infrastructure, and Ollama wants the same for open-weight models like GLM, Nemotron, DeepSeek, Kimi, and MiniMax, models a developer can now run "as easy as running any other piece of software," the company says, without an API key or dedicated server hardware.

The funding lands as inference infrastructure has become one of AI's hottest spending categories. Groq raised $750 million earlier this year to scale its inference chip business, as we covered. Open-weight models are increasingly the base layer for production inference outside major labs' clouds, including automated repository triage that would otherwise depend on paid API calls (a use case local LLMs already handle at scale).

Ollama did not disclose its valuation or say how the funds will be split between headcount, cloud infrastructure, and go-to-market spending.