Pricing
Free
Get started with Ollama
- Automate tasks with open models
- Keep your data private
- Run models on your hardware
- Run 1 cloud model at a time
- 100+ models
- CLI, API, and desktop apps
- 40,000+ community integrations
- Unlimited public models
Pro
Solve harder tasks, faster
- Run 3 cloud models at a time
- Faster responses from cloud hardware
- Larger models for challenging tasks
- 3 private models
- 3 collaborators per model
Max
For your most demanding work
- Run 5+ cloud models at a time
- High usage capacity for tasks that require heavy throughput
- 5 private models
- 5 collaborators per model
Additional usage available. Contact us
Power any task, big or small, with open models.
Your data remains yours.
Frequently asked questions
-
Does Ollama log any prompt or response data?
Ollama does not record, log or train on any prompt or response data.
-
What are cloud models?
Cloud models run on datacenter infrastructure, giving you faster responses and access to larger models. Your plan determines how many can run at the same time.
-
What are the usage limits?
Running models on your own hardware is always unlimited. Ollama's cloud infrastructure has limits that vary by plan:
- Free: Light usage—chat, quick questions, trying out models
- Pro: Day-to-day work—RAG, document analysis, regular coding assistance
- Max: Heavy, sustained usage—coding agents, batch processing, automation pipelines
Limits are designed to prevent abuse, not slow you down. If you run into issues, email us.
-
Is my data encrypted?
Yes, all cloud requests are encrypted in transit. We don't store your prompts or outputs.
-
Can I use Ollama in air-gapped environments?
Yes, Ollama runs fully offline on your own hardware. Cloud features are optional.
-
How does billing work?
Monthly subscription, billed at the start of each cycle. You can cancel anytime.
-
Do you offer team or enterprise plans?
Team and enterprise plans are coming soon. Contact us at [email protected] for more details.
-
Which cloud models are available?
See the full list of cloud-enabled models here.









