Ollama is a way to run open AI models on your own computer or in the cloud. The site says it runs on macOS, Windows and Linux, and describes it as the most popular way to build with open models. It also states that prompts are never stored or trained on, which it describes as zero data retention.
Models can run locally on your own hardware, and the pricing page says running them that way is always unlimited. Cloud models are also available and are billed per token, with prices listed per million tokens for each model. The model list includes options such as gpt-oss, Gemma, DeepSeek, Kimi and GLM, and the site says cloud models that support tool calling are tested with agent workflows before release.
It is aimed at developers and teams who want to build with open models, including people who run several agents at once. The Pay as you go tier is free to join, with usage charged per request and 3 concurrent requests. Pro costs $20 per month, Max $100 per month, and Team $500 per month with shared usage and centralised administration. Enterprise pricing is custom.
Usage is measured in tokens at each model's rates, and paid plans include a set dollar amount of usage credits each month. Concurrency limits control how many requests run at once, and requests beyond the limit are queued. The site says cloud models are hosted primarily in the United States, with possible routing to Europe for extra capacity.
