Ollama

Ollama is free, open-source software that runs AI language models on your own computer or server. Prompts and answers stay on your machine, so it suits organisations that cannot put client, case or donor data into ChatGPT. The models are smaller than the big cloud assistants and setup is more technical, so it is best with some IT help. Paid cloud plans also exist, but those send data to Ollama's servers.

Pricing: Running models locally is free (MIT licence). Optional cloud plans: Pro US$20 per month, Max US$100 per month, Team US$500 per month with shared usage credits. Cloud models run on Ollama's servers, so do not use them for data that must stay on your machine.

The product runs AI models locally. Quality depends on the model you download and the hardware you have.

Ollama states it does not collect, store, transmit or have access to prompts, responses or content you process locally. It collects limited metadata such as app version and request counts. Optional cloud models process content on Ollama's servers and, per its privacy page, do not store it beyond the time needed; Ollama says it does not train on user inputs or outputs. For genuinely sensitive data, use local models only and keep your device encrypted and patched, since the security of your own computer now matters.

Prices and offers confirmed on the vendor's own pages. Prices and nonprofit programs change often, so check with Ollama before you commit. Rankings on Nonprofit Tools are never for sale.