API & Platform
Ollama
Ollama lets you run models on your own computer. It also offers cloud models billed per token.
Is Ollama free?
Yes for local use. The $0 plan lists running models locally with no subscription, while cloud models are billed per token.
From the official pages listed under Sources. Last checked 2026-11-19.
Free plan
| What | What the official page says | Source (last checked 2026-11-19) |
|---|---|---|
| $0 plan (local models; cloud models billed per token) | $0 — Pay per token for every cloud model. — Run models locally — No subscription — 3 concurrent requests — No service fees | ollama.com |
Paid plans
| What | What the official page says | Source (last checked 2026-11-19) |
|---|---|---|
| Pro | $20 / mo. or $200/yr billed annually — $60 of usage credits per month | ollama.com |
| Max | $100 / mo. — $300 of usage credits per month | ollama.com |
| Team | $500 / mo. — Unlimited users; $1,000 of usage credits per month, shared across the team | ollama.com |
Fit and limitations
Editorial judgement, based only on the facts above
May suit
- People who want to run models locally with no subscription.
May not suit
- People who want cloud models without paying per token. On the $0 plan cloud models are billed per token.
What people say (public discussions, read on 17 November 2026)
This is a summary of other people's public posts, not a review by the site owner.
The discussions were posted between 30 July 2025 and 16 April 2026. Prices and limits they mention may have changed since.
Some posters like the one-command model download and the easy switching between models, and one says the desktop app is a good way to set up a non-technical friend with a local model.
perfect for setting up a non-technical friend or family member
(Hacker News thread, 30 July 2025)From: Hacker News thread, 16 April 2026; Hacker News thread, 30 July 2025
Others argue that Ollama is a wrapper around llama.cpp and that llama.cpp or other tools now cover the same ground. One poster says they moved to llama.cpp after frustrating experiences.
On the cloud service, some posters question the turn toward paid hosted models and how much privacy it offers. One poster who says they pay for the Pro plan likes testing several open models quickly without downloading them.
From: Hacker News thread, 5 August 2025; Hacker News thread, 16 April 2026
Posters disagree about the desktop app: some see a shift from developers to regular users that they do not want, others say it makes sense and widens the audience.
Discussions that were read
- The local LLM ecosystem doesn’t need Ollama (Hacker News, posted 16 April 2026; read 17 November 2026)
- Ollama Turbo (Hacker News, posted 5 August 2025; read 17 November 2026)
- Ollama's new app (Hacker News, posted 30 July 2025; read 17 November 2026)
Posters are not named. Posts are paraphrased, with at most one short quote per discussion. Reports about billing, privacy or safety are what some posters say, not facts confirmed here.
Questions people ask
Is Ollama free?
The pricing page lists a $0 plan that includes running models locally, no subscription and no service fees. On that plan you pay per token for every cloud model.
Source: ollama.com/pricing
What do the paid plans cost?
Pro is listed at $20 a month, or $200 a year billed annually, with $60 of usage credits per month. Max is $100 a month with $300 of credits, and Team is $500 a month.
Source: ollama.com/pricing
How many requests can run at once on the $0 plan?
The page lists 3 concurrent requests on the $0 plan.
Source: ollama.com/pricing