Ollama, the wildly popular open-source tool for running AI models locally, just dropped a huge funding announcement: a $65 million Series B led by Theory Ventures. This brings its total raise to $88 million, following a $15 million Series A led by Benchmark. Ollama’s magic is making complex open-weight models run on any PC in minutes—no PhD required. The team behind it? The same folks who gave us Docker Desktop. They’re basically doing for AI what Docker did for cloud apps: abstracting away the headaches. With 8.9 million monthly dev users and 85% of Fortune 500 companies onboard, all run by just 14 employees, this is a lean machine. Founders Jeff Morgan and Michael Chiang say the real inflection point came in January when open models suddenly became capable of real agentic tasks like coding. Why it matters: This funding signals that open models are no longer just a curiosity—they’re becoming a serious, cost-effective alternative to closed models like Anthropic for enterprises and startups looking to save on inference costs.