Ollama: $88 million in funding for local AI — 8.9 million developers and Docker's founder among investors

Ollama, the most widely adopted platform for running open AI models locally, announced $88 million in funding on July 9, 2026. Local inference means running models on your own hardware — without API keys, per-token billing, or sending data to external services, which is critical for privacy and regulated industries.

Who is behind the round?

The investment is led by Benchmark (partner Peter Fenton) and Theory Ventures (Tomasz Tunguz), joined by Y Combinator, 8VC, Garage Capital, and — symbolically most interesting — Solomon Hykes, Docker’s founder. Ollama’s founders, Jeff Morgan and Michael Chiang, also come from Docker (they worked on Kitematic), and Hykes’s investment confirms the thesis that Ollama plays for AI models the role Docker played for applications: a standardized, portable runtime.

Numbers that justify the investment

Ollama today serves 8.9 million developers and is used by 85% of Fortune 500 companies — a figure that shifts local AI from a hobbyist niche into enterprise mainstream. The Ollama cloud, a hybrid service for models too large for local hardware, doubles its processed token volume every month. For comparison, a year ago the platform was measured in hundreds of thousands of users.

What’s next?

The founders announce “a personal computer moment for AI”: hybrid inference that transparently combines local and cloud execution, plus first-day support for new models (GLM, Nemotron, DeepSeek). The funding arrives at a moment when open models — Qwen, DeepSeek, Llama, OLMo — are catching up with closed models in capability, making the infrastructure to run them a strategic layer. With $88 million, Ollama now has the capital to maintain that layer.