Last Updated: 9 July 2026
Ollama, an open-source platform that lets developers run AI models on their own computers with a single command, has raised $65 million in Series B funding. Co-founded by Jeffrey Morgan and Michael Chiang, the company says 8.9 million developers now use its platform, the largest developer network in the open model ecosystem. That number has doubled since January 2026.
Open-source AI models are catching up to proprietary ones. According to a 2026 market report, the open-source AI model market is projected to grow from $19.05 billion in 2025 to $23.08 billion in 2026, at a compound annual growth rate of 21.1%. Forbes reported in April 2026 that the performance gap between the best closed and open models has narrowed to just a few percentage points. Organizations are drawn to open models for practical reasons: lower costs, stronger data privacy, and less dependence on any single vendor's pricing.
Theory Ventures led the round, with participation from Benchmark, 8VC, Y Combinator, Pace Capital, 49 Palms, GTMFund, and other investors and angels. Total funding now stands at $88 million. Benchmark led the earlier Series A, and its General Partner Peter Fenton joined the board.
Several forces are pushing companies toward open-source AI. One is cost. Running a trained AI model to produce results, a process called inference, has become a major expense. Some companies have exhausted their annual AI budgets within months because usage-based pricing makes spending difficult to forecast. Open models sidestep this problem by running on a company's own hardware, which avoids expensive per-use fees. Regulators are also paying closer attention to where AI models come from and who controls them, making open-source options especially relevant for industries that handle sensitive data.
Ollama serves as a distribution and early-access partner for model labs including Meta, Google DeepMind, Mistral, and MiniMax, and for hardware companies such as NVIDIA, Intel, AMD, and Qualcomm. Users get access to new models on the day they are released, along with performance gains from deep hardware integration. More than 67,000 community-built integrations exist on GitHub, connecting open models to coding agents, personal assistants, document workflows, and other tools developers already use.
"Open-weight models will generate the supermajority of tokens within the next 18 to 24 months. The Ollama team saw this inevitability early — and understood the needs it would create for developers looking to discover, experiment with, and ultimately run these models at scale. They've already attracted millions of builders, and this is still just the beginning."
Peter Fenton, General Partner at Benchmark
The new capital will go toward product development, growing the open-source developer community, expanding cloud infrastructure, and hiring. A key priority is expanding the cloud service, which lets developers who need more computing power run the same models in Ollama's cloud without changing their code, creating a new account, or learning a different interface.
Cloud usage is growing quickly. According to a company blog post, token volume in Ollama's cloud has more than doubled every month on average. Ollama is also building what it calls "hybrid inference," a setup where models run locally or in the cloud depending on the task, with no manual switching required.
"Every era of computing has had a platform layer that everything else plugs into. As open models close the gap for most real work, the platform where AI runs becomes one of the most valuable positions in software. That market is expanding fast: AI is making software creation more accessible, and every professional can now build their own workflows and tools. Jeff and Michael recognized that shift early on, and built Ollama to meet those people where they already work."
Tomasz Tunguz, General Partner at Theory Ventures
Jeffrey Morgan (CEO) and Michael Chiang founded Ollama after meeting in college, where they built their first company, Kitematic. That tool simplified the use of Docker, a popular piece of software for packaging and running applications. Docker acquired Kitematic in 2015, and the founders' work became Docker Desktop, now used by over ten million developers. With Ollama, they took the same approach: make something powerful simpler to use.
At its simplest, Ollama is an application developers download to their computer. One command is all it takes to run an open AI model, meaning a model whose design is publicly available and free to use, directly on the user's own hardware. Because data stays on the user's machine, the tool is especially useful in regulated industries such as healthcare, finance, and government. Ollama does not train on user data. The platform is used within 85 percent of the Fortune 500.
"Open models should be easy to run, easy to build with, and available wherever people need them — on your own machine, in the cloud, or both. Ollama started as an open-source project, and has since grown into a community of millions of developers. Everything we do next is in service of that community and their best work. The future of AI is open models running everywhere work gets done — and Ollama is here to power this shift."
Jeffrey Morgan, CEO and Co-Founder of Ollama
Theory Ventures, founded by Tomasz Tunguz, led the Series B. The firm focuses on early-stage software companies working with AI and data. It raised a $450 million second fund in late 2024, following a $230 million debut fund in 2023. Benchmark, which led the Series A, is an early-stage venture firm known for investments in companies such as Uber, eBay, and Twitter. Its General Partner Peter Fenton sits on Ollama's board.
Other participants in the round include 8VC, Y Combinator, Pace Capital, 49 Palms, and GTMFund. Angel investors include Solomon Hykes, the founder of Docker; Aaron Katz, CEO of ClickHouse; Spencer Kimball, co-creator of GIMP and co-founder of Cockroach Labs; Quinn Slack, CEO of Amp; and Marianna Tessel, who serves on the board of directors at Cisco.
