Md Sarwar
Ollama is a platform for running and building with open AI models, either on a user’s own computer or on Ollama’s cloud. It is available for macOS, Windows, and Linux, and the site states that prompts are never stored or trained on.
The documentation describes uses such as connecting open models to desktop apps and coding agents, and building applications that call cloud models through Ollama, OpenAI, or Anthropic-compatible APIs.
Python and JavaScript libraries are also provided. Running models on your own hardware is described as always unlimited. Cloud models are billed per token under a pay-as-you-go plan or covered by monthly usage credits on Pro, Max, Team, and Enterprise subscriptions.
How It Works
Users download Ollama and run a model on their own computer, or choose cloud models that run without a local GPU. Applications can reach cloud models through Ollama’s API or through OpenAI- and Anthropic-compatible clients.
Cloud usage is metered in tokens at each model’s published rate, drawing first on included plan credits and then on any purchased credit balance.
Why Users Choose It
- Local use of models on own hardware has no usage cap and 100% free.
- Prompt and response data is stated to never be logged or trained on
- Offers both local and cloud model options on one platform
- Compatible with OpenAI and Anthropic API clients
- Pricing ranges from pay-per-token to team and enterprise plans
Core Features
- Run open AI models locally on macOS, Windows, and Linux
- Cloud models billed per token, with no local GPU needed
- Tool calling support for cloud models trained for it
- Documented capabilities including streaming, thinking, structured outputs, vision, embeddings, speech, and web search
- API with OpenAI- and Anthropic-compatible clients
- Python and JavaScript libraries
- Integrations with desktop apps and coding agents
- Team and Enterprise administration features such as centralized billing and model access controls
Use Cases
- Running open models privately on a personal computer
- Powering coding agents and desktop apps with open models
- Building applications that call cloud-hosted open models
- Running multiple agents at once on higher-tier plans
- Sharing model usage and billing across a team
Pros & Cons
Pricing
Pricing model: Freemium
| Plan | Monthly | Annual | Limits | Notes |
|---|---|---|---|---|
| Pay as you go | $0 | Not stated | Pay per token for every cloud model; 3 concurrent requests | Includes running models locally, no subscription, and no service fees |
| Pro | $20 / mo. | $200/yr billed annually ($16.67 / mo.) | $60 of usage credits per month; 3 concurrent requests | Access to larger pro models; fast mode listed as coming soon |
| Max | $100 / mo. | Not stated | $300 of usage credits per month; 10 concurrent requests | Early access to the newest models; aimed at power users running multiple agents |
| Team (early access) | $500 / mo. | Not stated | $1,000 of usage credits per month, shared across the team; unlimited users; 10 concurrent requests | Centralized billing and administration and priority support; shared projects, skills, and instructions listed as coming soon |
| Enterprise | Custom | Custom | Volume usage pricing | Adds model access controls, cost budgets for users and API keys, a private Slack channel with dedicated support, and custom security questionnaires |
Cloud model prices are listed per million tokens (input, cached input, output). Off-peak pricing applies for some models outside 12:00 and 18:00 UTC on weekdays and all day on weekends. Included usage resets monthly and does not roll over. Verify current pricing on the official Ollama pricing page.
