Ollama Raises $65M to Grow Open-Source AI Platform

The platform that's quietly become the default way developers run open-source AI models on their own machines just raised serious growth capital. Ollama announced a $65 million Series B funding round led by Theory Ventures, bringing the company's total funding to $88 million, with participation from Benchmark, 8VC, Y Combinator, and Pace Capital, among others, according to SiliconANGLE's reporting on the raise.

The traction numbers explain the investor enthusiasm. Ollama's developer base has doubled since January 2026 alone, reaching 8.9 million monthly active developers, with the platform adding nearly one million new installs every week, according to HPCwire's coverage of the announcement. The company, run by a lean 14-person team, is now used within 85% of the Fortune 500, including customers in deeply regulated sectors like healthcare, finance, and government, a level of enterprise penetration that's unusual for a tool this technically oriented, one worth understanding alongside our broader coverage of what a large language model actually requires to run outside a cloud environment.

Ollama's core pitch is simplicity. With a single command, developers can download and run an open AI model locally on their own hardware, then seamlessly scale to more powerful models in Ollama's cloud without changing their code or workflow at all. That flexibility matters because it lets developers avoid per-token cloud costs for quick, low-latency tasks while still having cloud scale available when a task genuinely needs it.

Benchmark general partner Peter Fenton, who sits on Ollama's board, framed the underlying market shift in stark terms, according to HPCwire's reporting. "Open-weight models will generate the supermajority of tokens within the next 18 to 24 months," he said. That prediction lines up with the broader momentum we've tracked around Chinese open-weight models gaining ground against closed U.S. competitors, a trend worth pairing with our DeepSeek statistics coverage for the fuller picture.

Ollama's growth isn't happening in isolation. The company provides early access to models from Meta, Google DeepMind, Mistral, and MiniMax, while working closely with hardware providers including Nvidia, Intel, AMD, and Qualcomm to optimize performance across different chip architectures. The platform supports more than 67,000 community-built integrations, giving it a genuinely broad developer ecosystem rather than a single flagship use case.

CEO and co-founder Jeffrey Morgan, whose previous startup Kitematic was acquired by Docker, described the company's trajectory in similarly expansive terms, according to SiliconANGLE's reporting. "Ollama started as an open-source project, and has since grown into a community of millions of developers," he said, a philosophy that connects directly to the growing enterprise interest in AI coding tools that don't require sending proprietary code to an external cloud provider.

I've advised companies on AI adoption for four years, and Ollama's traction is worth understanding even if your organization has never touched an open-weight model directly. For businesses in regulated industries specifically, healthcare, finance, and government among Ollama's stated customer base, the ability to run AI models entirely on local infrastructure without sending data to an external cloud is a genuinely different value proposition than subscribing to a hosted AI service, one built around data control rather than raw capability.

For technical teams evaluating AI infrastructure options, Ollama's rapid growth and Fortune 500 penetration is a strong signal that local and hybrid model deployment is becoming a mainstream enterprise strategy, not a niche developer preference.

Ollama raised $65 million in Series B funding, led by Theory Ventures, bringing its total funding to $88 million. The platform's developer base doubled since January to 8.9 million, and it's now used within 85% of the Fortune 500, including regulated industries like healthcare and finance. Ollama lets developers run open AI models locally with a single command, then scale to the cloud without changing their workflow.