Ollama Raises $65M Series B for Open-Source AI Infrastructure

Artificial intelligence spent the last few years chasing bigger models, bigger GPU clusters, and bigger cloud bills. Ollama is betting the next chapter will be written closer to the developer, where model access, local execution, and infrastructure control actually determine what gets shipped.

The open-source AI platform announced a $65M Series B led by Theory Ventures, bringing total funding to $88M. Ollama has grown to nearly 8.9M monthly developers, with usage across 85% of the Fortune 500, turning what started as developer tooling into a serious enterprise AI infrastructure signal.

What Happened

Ollama raised $65M in Series B funding led by Theory Ventures, with participation from Benchmark, 8VC, Y Combinator, Pace Capital, 49 Palms, GTMFund, and other investors and angels. The round follows a $15M Series A led by Benchmark, with Peter Fenton joining the board, and brings Ollama's total funding to $88M.

Founded by Jeffrey Morgan and Michael Chiang, Ollama gives developers a simpler way to run open-weight large language models locally and through hosted infrastructure when workloads require more scale. The company's GitHub repository has become one of the most visible developer entry points for local AI workflows.

Why This Matters

The funding matters because AI deployment is moving from model-choice theater into infrastructure reality. Enterprises are asking where data lives, how inference costs behave at scale, how easily teams can swap models, and whether developers can build without waiting for a platform committee to bless every experiment.

Ollama sits directly inside that shift. Open-weight models are models whose trained parameters are available for developers to run outside a closed API-only environment, and Ollama makes those models easier to bring into practical workflows. That means the company benefits from a broader market shift toward flexible AI stacks, where local execution, hosted open models, and proprietary APIs can coexist instead of forcing a single deployment philosophy.

Market Context

Open-source AI has evolved from a developer enthusiasm loop into a legitimate enterprise infrastructure category. The market is no longer asking only which model tops the benchmark charts. It is asking which tools help teams control privacy, latency, cost, model choice, and operational complexity without turning every deployment into a research project.

That is why developer adoption matters here. Nearly 8.9M monthly developers is not just a vanity metric when the product sits inside the workflow of people deciding what infrastructure becomes standard. Developers tend to adopt tools that save time under pressure, and once a tool becomes part of the daily workflow, replacement costs are measured in interrupted work, not just software spend.

Competitive Landscape

Ollama is not competing in the same lane as every foundation model company. Model labs compete to produce more capable systems, cloud providers compete to sell compute, and application companies compete for end-user attention. Ollama's lane is the developer experience around discovering, running, and scaling open models.

That positioning matters because Ollama does not need a single open model to emerge as the winner. As the open-model ecosystem expands, the need for simple local and hosted deployment grows alongside it. The company becomes more valuable if developers continue demanding choice, and choice is exactly what enterprises begin asking for once AI moves from experimentation into recurring workloads.

What This Signals

Theory Ventures leading the round points to growing investor conviction around AI infrastructure companies that make the broader ecosystem usable. The glamorous part of AI may still be model releases, but durable value often accumulates around the products that make those releases deployable, manageable, and reliable enough for serious production work.

Benchmark's earlier backing reinforces the same pattern. Infrastructure companies can look less dramatic than application-layer startups until the market realizes the infrastructure is what everyone else depends on. Docker demonstrated that lesson in cloud software, and Ollama is now carrying a version of it into open-source AI.

The Bigger Industry Shift

AI buyers are becoming more practical. They still care about capability, but they also care about whether teams can deploy securely, manage costs, avoid unnecessary vendor lock-in, and keep infrastructure flexible as models evolve. Those questions increasingly shape purchasing decisions because the hard part is no longer proving AI can be useful. The hard part is making it useful repeatedly without losing control of the stack.

Ollama's Series B is ultimately a bet that developer productivity remains one of the strongest leading indicators of enterprise adoption. History tends to reward companies that make difficult technology feel ordinary, and AI may reward the platforms that remove enough complexity for open models to become everyday infrastructure instead of a specialist exercise.