Software · 9 July 2026 · 3 min read
The "Docker" of Local AI: Ollama Raises $65 Million to Win Over 9 Million Developers
In brief: Ollama has announced a $65 million Series B funding round, bringing its total valuation support to $88 million. Founded by the creators behind Docker Desktop, the startup enables developers to run open-weight AI models locally with extreme ease. It currently boasts nearly 9 million monthly active users and is utilized within 85% of Fortune 500 companies.
by Team Mocchi's
A Quiet Revolution on the Developer's Desktop
Over the past three years, the AI gold rush has heavily focused on massive, cloud-hosted proprietary models. In parallel, however, a quiet yet powerful ecosystem has emerged: local AI and open-weight models (such as Meta's Llama or Google's Gemma). Running a language model on one's own hardware offers massive benefits in terms of privacy, latency, and cost control. Historically, though, doing so required complex hardware configurations, CUDA drivers, and manual dependency management.
Ollama stepped into this space by removing these barriers. With a single terminal command, Ollama allows developers to download and run complex models directly on their personal computers, automatically optimizing resource allocation across CPUs and GPUs (including Apple Silicon).
Exponential Growth by the Numbers
The success of this approach is validated by impressive financial and adoption metrics. As reported by TechCrunch, Ollama has just closed a $65 million Series B funding round led by Theory Ventures, with participation from previous investor Benchmark. This brings the startup's total funding to $88 million.
The adoption metrics are even more striking. Operating with a lean team of just 14 employees, Ollama has reached nearly 9 million monthly active developers, amassing over 176,000 stars on GitHub. The tool has also deeply penetrated the enterprise landscape, finding a home in 85% of Fortune 500 companies. This explosive growth proves that testing and integrating AI securely and locally has become a top priority for both individual developers and large enterprises alike.
From Cloud to Terminal: The Docker Legacy
Ollama's extraordinary usability is no accident; it is the direct result of its founders' past experience. Jeff Morgan and Michael Chiang previously built Kitematic, a startup acquired by Docker that subsequently laid the groundwork for Docker Desktop.
Just as Docker revolutionized cloud computing by abstracting operating system complexities to let software run anywhere via containers, Ollama is applying the exact same philosophy to artificial intelligence. Before Ollama, models released by research labs were primarily designed for academics and data scientists. Ollama made them accessible to any software developer by standardizing distribution formats and abstracting the intricacies of the underlying hardware.
Monetization Strategy: From Local to the "Neocloud"
Despite its deep open-source roots, Ollama has established a clear monetization roadmap. Alongside its free, local-first version, the company offers a proprietary cloud service known as the "neocloud." Through several subscription tiers scaling up to $100 per month, developers can access larger and more complex models that would otherwise be too heavy to host on a standard laptop.
The key differentiator of this offering lies in its pricing model: unlike traditional API providers (such as OpenAI or Anthropic), which charge based on the number of processed tokens, Ollama bills customers based on actual GPU compute time. This approach ensures highly predictable operational costs for companies integrating AI within their development pipelines and production environments.
Mocchi's Take
Ollama's rise and the success of its latest funding round confirm that AI is no longer a monopoly of large US cloud providers. For Italian enterprises, adopting local AI represents an unprecedented strategic opportunity to bypass strict GDPR compliance hurdles and safeguard proprietary intellectual property. As a software agency, we believe that tools like Ollama are finally democratizing the development of custom software solutions, allowing us to test and prototype intelligent agents locally before scaling them onto private infrastructures with incredibly low operational and inference costs.