Ollama: All Aboard Open Models
Ollama: Empowering the Open Model Revolution
The journey began years ago in college, where Michael and I co-founded Kitematic. Our goal back then was simple: make running Docker dead-simple. That effort eventually evolved into Docker Desktop, a tool launched in 2016 that has since been adopted by over ten million developers globally.
A decade later, we have reunited to build Ollama—the most streamlined path for developers to deploy and utilize open-source models.
To date, Ollama has grown beyond our expectations, becoming the premier platform for open models. We now serve 8.9 million developers who share our conviction: AI should be something you own, build upon, and control.
The "Personal Computer" Moment for AI
We are witnessing a shift similar to the birth of the PC. Just as the personal computer migrated computing from massive, locked-away mainframes to the individual's desk, we are doing the same for AI.
"The personal computer took the machine out of the mainframe room and put it on your desk — yours to own, customize, and build with."
A few years ago, high-performance open models began to emerge. While they were free to download, the barrier to entry was high; getting them to actually work was a struggle. The raw power existed, but it wasn't accessible to developers who wanted a seamless experience similar to proprietary APIs.
To bridge this gap, we created Ollama. It is an application that allows you to launch the latest open models with a single command and integrate them via a straightforward API.
With Ollama, deploying an open model is now as easy as installing any other software. There are no permission hurdles, no API keys, and no requirements for massive server farms.
The Three Pillars of Ollama
We designed Ollama around three core values that resonate with our community:
| Principle | Description | Benefit |
|---|---|---|
| Ownership | You control the models your apps rely on. | No vendor lock-in; total freedom to modify. |
| Cost | Models run on your own local hardware. | No runaway per-token billing; iterate for free. |
| Privacy | Data remains on your local machine. | Maximum security; trust is maintained even when scaling. |
In mathematical terms, we believe the value of AI is maximized when:
Scaling for the Enterprise
Open models are no longer just for research or hobbyists. They are being integrated into the world's most influential organizations—in fact, 85% of the Fortune 500 now utilize Ollama.
What began as the simple thrill of running a model locally has transformed into a mission to solve complex problems that were previously the sole domain of proprietary systems. To support this, Ollama's cloud provides a seamless way to access elite open models, including:
DeepSeekNemotronGLMKimiMiniMax- ...and many others.
Our cloud infrastructure is experiencing explosive growth, with token volume more than doubling every single month.
All In on Open: The Path Forward
We are thrilled to announce that Ollama has secured $88M in funding. This round was led by:
- Peter Fenton (Benchmark)
- Tomasz Tunguz (Theory Ventures)
- Alex Kolicich (8VC)
We are also supported by an incredible group of strategic investors:
- Solomon Hykes (Founder of Docker)
- Aaron Katz (CEO of ClickHouse)
- Spencer Kimball (Co-founder of Cockroach Labs / GIMP co-creator)
- Quinn Slack (CEO of Amp)
- Marianna Tessel (Cisco Board)
- Michael Montano (Former Twitter Head of Engineering)
- Y Combinator, Garage Capital, Pace Capital, 49 Palms, GTMFund, and various angel investors.
Our Roadmap
Ollama is positioned at the heart of the open model ecosystem. We are committed to pushing the boundaries further through:
- Seamless Hybrid Inference: Blending local and cloud power.
- Instant Support: Adding new open models the very day they are released.
- Privacy-First Cloud: Enabling teams to scale without sacrificing ownership.
We owe a debt of gratitude to the researchers and teams releasing these powerful models. Our mission is to ensure their work is effortless to deploy and remains open for all.
We have bet on the idea that "open and easy" wins before, and we are doing it again.
Ready to start? Download Ollama today and begin building:
curl -fsSL https://ollama.com/install.sh | sh
Thank you to every developer who has run a model, reported a bug, created an integration, or shipped a product using Ollama.