Ollama, the open source AI developer tool that helps people run AI on their PCs, has raised $65M and grown to nearly 9M users, TechCrunch reports. The same piece of evidence that matters to engineers and investors is also visible in the project’s traction: Ollama has amassed 176,000 stars and nearly 17,000 forks on GitHub.

Put simply, this is not “AI toy demos” growth. Ollama is building a developer workflow around local execution, and it is doing it at scale. Benchmarks backing the approach and the community momentum on GitHub tell a consistent story: developers are choosing a tool that runs where they already work, instead of treating AI like something that only lives in the cloud.

So why does $65M matter here, beyond the headline number? Because open source AI tools are starting to look more like infrastructure than experiments. When a project like Ollama can plausibly reach nearly 9M users while posting 176,000 stars and nearly 17,000 forks, it means adoption is widening across the “build and test” lifecycle, not just the “try it once” curiosity stage.

Ollama’s GitHub metrics underline that difference. Stars are a rough signal of interest, but forks are a more active signal that developers want to modify, extend, or integrate the system into their own setups. With nearly 17,000 forks, the ecosystem is not only watching. It is iterating. That typically attracts more contributions, which can improve features, stability, and compatibility. It can also accelerate the pace at which new model support and local runtime improvements arrive, since an open ecosystem can parallelize work across teams.

There is also a business incentive hiding in plain sight: local AI execution can change cost curves and operational risk assumptions. When developers can run AI on their PCs, the “per request” cloud tax is not the only default option. That can push teams to prototype faster, test more variations, and build internal tooling without constantly waiting on external capacity. Even when cloud use remains necessary for production scale, local tools often become the earliest stage of development, where iteration speed is the competitive advantage.

Regulatory and compliance pressures add another layer to the interest. In many environments, organizations want clearer control over where data goes and how it is processed. Local execution can support privacy and data governance strategies that do not require sending every experiment to a third-party service. TechCrunch did not spell out a specific regulatory framework in this report, but the category of “run AI locally” tends to align with common compliance goals: minimize data sharing, reduce exposure surfaces, and maintain more direct control over compute.

For boards and investors, the second-order question is whether Ollama is only a tool or whether it is becoming a platform. Nearly 9M users suggests broad reach, but the more important part is how developers connect it into their daily workflow. A developer tool with heavy community usage can become a default choice, which in turn can shape which models get integrated first, how integrations are built, and which ecosystems become “compatible by default.” Capital then becomes a mechanism to harden reliability, improve onboarding, and expand the surfaces where enterprises and power users can adopt.

For peers who are also funding AI infrastructure, the implication is uncomfortable in a good way: local AI is not fringe anymore. Ollama’s $65M raise combined with its measurable community footprint suggests incumbents and challengers alike should assume PC based AI tooling will keep absorbing developer attention. If you are an executive evaluating AI strategy, the strategic stake is clear. The winners may not only be the largest model providers. They may be the tools that turn AI from an external service into a normal, repeatable development habit.