Ollama

Ollama provides software for developers to run open models on their own computers or access larger models through Ollama’s cloud. Its app and API let developers build AI applications, while integrations connect models to coding tools including Claude Code and Codex. Local execution keeps data on the user’s machine; cloud execution provides access to models that would not fit on a personal computer. Modelfiles let developers create and share customized models, specifying prompts, runtime parameters and adapters.

Founded by Jeffrey Morgan, its CEO, and Michael Chiang, Ollama follows their earlier work on Kitematic, which Docker acquired in 2015 and which became part of their work on Docker Desktop. Ollama’s engineering focuses on making models portable and practical to run. Its multimodal engine uses the GGML tensor library, with self-contained model implementations and model-specific projection layers. It also caches processed images for reuse in later prompts.

In July 2026, the company reported 8.9 million monthly active developers and more than 67,000 community-built integrations. That month, Ollama announced a $65 million Series B led by Theory Ventures, bringing total funding to $88 million. The financing supports product development, its open-source developer community, expansion of cloud compute capacity and hiring.