We may earn a commission if you buy through links on this page. It never changes a score or a ranking. How we make money
Paid plans start at $16.67/mo. There's also a free plan.
Ollama's local runtime is open-source under the MIT license, so you can start without a software subscription and run models on hardware you provide. Hosted cloud use has a Pay as you go plan with a $0 subscription, but it is billed per token. For teams that want monthly credits and higher request concurrency, paid plans run from Pro at $20 per month to Team at $500. The right budget depends on whether inference stays local or moves to hosted models.
01Every plan, line by line
| Plan | Billed annually | Billed monthly | Built for |
|---|---|---|---|
| Pay as you go | — | Quote | Local models and occasional cloud use |
| Pro | $16.67 | $20 | Short, well-defined daily tasks |
| Max | — | $100 | Power users running multiple agents |
| Team | — | $500 | Teams sharing usage and administration |
| Enterprise | — | Quote | Larger teams with volume usage pricing |
Add-ons and overages
- Usage credits can be purchased; prices are not published here.
- Usage beyond Team's included credits is billed as you go from a shared team balance.
02What the plan grid doesn't tell you
Token usage
Pay as you go charges per token. The vendor doesn't publish token rates here, so get current rates and model-specific billing details before forecasting cloud spend.
Credit overages
Team usage beyond its included $1,000 in monthly shared credits is billed as you go from the shared team balance.
Extra credits
Additional usage credits can be purchased, but their prices are not published here. Ask for current rates when estimating production workloads.
Local hardware
Self-hosting avoids a software subscription, but you supply the machines and GPUs needed to run your chosen models.
03Which plan should you buy?
| If you are… | Get | Why |
|---|---|---|
| Local development or occasional cloud | Pay as you go | $0 subscription, three concurrent requests and local model use; hosted usage is billed per token. |
| Individual with regular cloud tasks | Pro | $20 monthly or $200 annually for $60 in monthly credits and three concurrent requests. |
| Team sharing hosted usage | Team | $500 monthly for unlimited users, $1,000 in shared credits, 10 concurrent requests and priority support. |
04Is Ollama worth the money?
Paying makes sense when hosted models, included monthly credits, higher concurrency or team administration are worth more than running inference on your own hardware. Pro is the entry point for an individual who expects regular cloud usage; Team bundles shared credits and priority support for multiple users. For local development, offline inference or workloads that fit available hardware, the open-source runtime can be enough. Model cloud costs against actual usage before choosing a plan, because token rates and extra-credit prices are not published here.
Ollama pricing FAQ
Is Ollama free to use?
The local runtime is open-source under the MIT license and has no subscription price. You still provide the hardware. Hosted Pay as you go has a $0 subscription, but cloud inference is billed per token, so it is not unlimited free hosted usage.
What is the cheapest Ollama cloud plan?
Pay as you go has a $0 subscription and includes three concurrent requests, with hosted usage billed per token. Pro costs $20 per month or $200 billed annually and includes $60 in monthly usage credits. Compare token rates and expected use before choosing.
How much does Ollama Pro cost annually?
The pricing page lists Pro at $20 per month or $200 billed annually, which is equivalent to $16.67 per month. Pro includes $60 in monthly usage credits and three concurrent requests. The page lists no annual prices for Max or Team.
What happens when Team credits run out?
Team includes $1,000 in monthly shared usage credits. Usage beyond those included credits is billed as you go from the shared team balance. The vendor also says usage credits can be purchased, but does not publish their prices here.