Ollama Raises $65M to Fuel Open-Source AI Tools for Developers

The open-source AI startup Ollama has secured $65 million in Series B funding led by Theory Ventures, bringing its total raised to $88 million, founder and CEO Jeff Morgan confirmed to TechCrunch. The round follows a $15 million Series A led by Benchmark's Peter Fenton. Since its launch in 2023, Ollama has become a go-to tool for developers looking to run open-weight AI models directly on their personal computers, with setup taking just minutes. The platform now boasts over 8.9 million monthly active developers and is used by 85% of Fortune 500 companies, all with a lean team of just 14 employees.

Ollama's appeal lies in its simplicity: developers can find, download, and run models locally, or access larger, more complex ones through its neocloud service, which offers subscription tiers from free to $100 per month. Usage is tracked by GPU time rather than token limits, a feature that sets it apart from many competitors. The company's founders, Jeff Morgan and Michael Chiang, previously built Docker Desktop - a tool that revolutionized cloud application deployment - before selling their earlier startup Kitematic to Docker. Morgan sees Ollama as doing for AI what Docker did for cloud computing: abstracting away hardware headaches.

The timing of the funding round aligns with a broader industry shift toward open-weight models, particularly since early 2024 when models like OpenClaw began handling complex coding tasks. Morgan notes that enterprises and AI startups are increasingly turning to affordable open models for daily operations, reserving closed models for specialized needs. Benchmark's Peter Fenton, who joined Ollama's board, says the debate between open and closed models is a false choice: both will thrive, but every company with high inference costs has a strong incentive to explore open-weight options. Despite some community pushback over the company's commercial cloud service, Morgan insists it's a natural extension of their mission to help developers access and use models, especially those too large to run locally.

Source: TechCrunch