Ollama, the open-source platform that lets developers run artificial intelligence models on their own computers, has raised $65 million in Series B funding and grown to nearly 9 million monthly users, signaling a major industry shift toward locally-hosted, cost-effective AI. The funding round, led by Theory Ventures, brings the company's total capital raised to $88 million since its 2023 launch.

The timing of this funding reflects a broader trend in the AI industry. As large language models (LLMs), which are AI systems trained on vast amounts of text to understand and generate human language, have become more capable, companies are increasingly looking for ways to reduce their inference costs, which are the expenses incurred when actually using these models to process information.

Why Are Enterprises Turning to Open-Source AI Models?

Ollama's growth reflects a fundamental shift in how organizations approach AI spending. Rather than relying exclusively on expensive proprietary models from companies like Anthropic, enterprises are now building hybrid strategies that use open-weight models, which are publicly available AI models with transparent architectures, for routine tasks and reserving costly closed models for specialized work.

"Every company with high inference expenses has a vital existential project pushing them to move to open-weight models," said Peter Fenton, Benchmark partner and board member at Ollama.

Peter Fenton, Partner at Benchmark

The shift accelerated around January 2026, when larger open models demonstrated they could handle complex agentic tasks, such as writing and debugging code, that were previously the domain of proprietary systems. This capability breakthrough convinced many organizations that open models could deliver real business value while significantly reducing costs.

How Does Ollama Make AI Models Accessible to Developers?

Ollama simplifies what was once a technically complex process. When open-source AI models first emerged in 2023, they were difficult to use and primarily designed for researchers rather than working programmers. Ollama abstracts away the hardware configuration challenges, allowing developers to download and run sophisticated AI models on their personal computers in minutes.

The platform operates similarly to Docker Desktop, a tool that revolutionized cloud development by making it easy to package and move applications across different computing environments. Ollama's founders, Jeff Morgan and Michael Chiang, previously built Kitematic, which Docker acquired. They applied the same philosophy to AI: making complex technology accessible to everyday developers.

• Free Desktop Tool: Developers can download and run open-source models locally without paying anything, with access to thousands of models through Ollama's repository.
• Cloud Compute Option: For models too large to run on personal computers, Ollama offers cloud-based compute through its neocloud service with subscription tiers ranging from free to $100 per month.
• Usage-Based Pricing: Unlike some competitors that charge based on the number of tokens (individual words or word fragments) processed, Ollama tracks costs based on GPU time, offering more predictable pricing for heavy users.

The company has achieved remarkable scale with minimal overhead. Ollama now serves over 8.9 million developers every month and operates in 85% of Fortune 500 companies, all with just 14 employees. The platform has accumulated 176,000 stars on GitHub, a code repository platform where developers share and collaborate on projects, and nearly 17,000 forks, indicating strong community adoption and contribution.

What Does This Funding Mean for the Open-Source AI Ecosystem?

Ollama's success is part of a broader wave of open-source AI infrastructure companies attracting venture capital investment. Other players in this space include Inferact, which created vLLM, a tool for efficient model inference, and RadixArk, which developed SGLang, another inference optimization framework. Even tiny startups are building their own open-source models from scratch, like Arcee.

Some community members initially expressed concern that Ollama's commercial cloud service might distract from its free open-source mission, citing what critics call "enshittification," a term describing the gradual degradation of free services as companies prioritize paid offerings. However, Ollama's leadership argues the cloud service is a natural extension of the platform's core mission.

"Nothing has changed for the core product that's free on the desktop. There's zero change to the premise that this is the place you can discover and run local models," said Peter Fenton, addressing concerns about the company's commercial direction.

Peter Fenton, Partner at Benchmark

Founder Jeff Morgan explained the rationale for the cloud offering: state-of-the-art, large open models are often too resource-intensive to run on personal computers, so Ollama's cloud service helps developers access compute power when needed without requiring them to build their own infrastructure.

The funding validates a thesis that has been gaining traction throughout 2026: open-source AI models are not replacing proprietary systems, but rather creating a complementary ecosystem where organizations use both. This hybrid approach allows companies to optimize costs while maintaining flexibility to use specialized proprietary models when necessary. For Ollama, it means the market for local and cloud-based open-source AI infrastructure is far larger than many initially anticipated.