AI comparison
Open-weights vs closed models
Open-weights models can be downloaded, run and adapted on your own hardware within their licence; closed models are used only through the provider's API.
| Aspect | Open weights | Closed model |
|---|---|---|
| Access | Download the weights | Call an API |
| Control | You choose version, hardware and changes | The provider decides |
| Adapting | Fine-tune freely within the licence | Only what the provider offers |
| Licence | Varies; read it before building on it | Terms of service |
Choose Open weights when
- You need to run it offline or on your own network
- You want to fine-tune or inspect the model
- You must keep one version for a long time
Choose Closed model when
- You want the strongest model without running servers
- You prefer to pay per use
- Your team has no time to operate models
Learn it hands-on
These free lessons teach it with commands to run in a lab and checked tasks. Read them whole with no account.
Related terms
- Open weights
- An open-weights model is one whose trained parameters are published for download, so you can run, inspect and fine-tune it on your own machines.
- Local model
- A local model is a language model that runs on hardware you control, a laptop, a server or your own cloud machine, rather than behind someone else's API.
- Fine-tuning
- Fine-tuning is training an existing model further on your own examples so it learns a task, style or format it did not do well before.