Ollama: The Leading Open Source AI Development Tool Raises $65M and Grows to Nearly 9 Million Users

Ollama, a popular open-source AI platform, has successfully raised $65 million in a Series B funding round led by Theory Ventures, as reported by CEO Jeff Morgan to TechCrunch.

This funding adds to a previous $15 million Series A round led by Peter Fenton of Benchmark, bringing the company’s total funding to $88 million.

Founded in 2023, Ollama enables developers to run open-weight AI models on their personal computers, offering a quick setup that takes only minutes. The tool has received widespread acclaim across various training platforms, videos, blogs, and social media, achieving 176,000 stars and nearly 17,000 forks on GitHub.

The platform allows developers to search for models and utilize larger, more complex ones hosted on its neocloud, available through several subscription levels ranging from free to $100 per month. It tracks usage based on GPU time rather than token limits.

If the emphasis on simplifying development for personal computers sounds familiar, it’s because both Morgan and co-founder Michael Chiang were instrumental in creating Docker Desktop. They joined Docker following its acquisition of their previous startup, Kitematic, which made cloud app portability smoother while minimizing complex hardware configuration challenges.

Essentially, Ollama aims to replicate the success Docker and Docker Desktop achieved for AI.

“Open models started to emerge in 2023, but they were tough to use,” Morgan pointed out. Initially designed for researchers rather than developers, they posed implementation challenges. “Now, three years post-launch, Ollama is leveraged by over 8.9 million developers monthly, present in 85% of the Fortune 500, while experiencing rapid growth,” he noted, all with a team of just 14 employees.

This professional history prompted Benchmark’s Peter Fenton to lead the previous funding round and join the board.

“What Jeff and Michael created with Docker is used by over 10 million developers daily. The ability to innovate and create a product that becomes ubiquitous among developers is astonishingly rare,” Fenton told TechCrunch.

Morgan and Fenton chose not to disclose the startup’s revenues or its latest valuation. However, Morgan highlighted that a significant moment for Ollama’s business came around January when OpenClaw gained traction. During that time, larger open models proved capable of completing tasks like coding. The rise of assistants like OpenClaw demonstrated the potential of open models to deliver real results.

Since then, extensive discussions have emerged in the industry regarding the increasing tendency of paying users—especially well-funded enterprises and emerging AI application-layer startups—to prefer more affordable open models while relying on closed models like Anthropic for only occasional use.

“I believe this is where much of the conversation misses the mark. It’s not simply a binary choice,” Fenton said regarding the debate between open and closed AI models. He believes there’s substantial room for both. However, organizations facing high inference costs—the expenses incurred when utilizing those models—are embarking on a “critical existential project” to transition to open-weight models, he asserted.

There is considerable evidence that startups and enterprises are already moving toward open models for their daily operations, which bodes well for Ollama’s cloud services.

Furthermore, Ollama showcases how AI is generating numerous new open-source projects that attract venture capital interest. Other open-source inference providers like Inferact, creator of vLLM, and RadixArk, developer of SGLang, are emerging. Alternatives like OpenClaw and NanoClaw are also developing, along with smaller startups such as Arcee that are building their own open models from scratch.

However, not all of Ollama’s supporters are satisfied with the company’s monetization strategy. About a year ago, various blog posts and social media comments expressed concern that its cloud business was shifting focus away from the beloved free project, citing Ollama as an example of the so-called “Enshittification” trend in developer tools.

Nevertheless, Morgan sees the cloud service as a natural evolution of Ollama’s open-source mission to help developers find and effectively use models. These advanced, large open models frequently require resources beyond what personal computers can provide. “So we thought, ‘Let’s facilitate access to that computing power,’” he explained.

Fenton added, “The core product that remains free on the desktop has not changed. The notion that this is the platform for discovering and running local models has remained intact.”