AI Providers
Configure your own AI vendors and models in Ember
AI Providers
Ember does not host models or bill for inference. Configure your own providers and API keys.
Supported integrations
Ember connects to common OpenAI-compatible endpoints and major vendor SDKs, including:
- OpenAI / compatible gateways
- Anthropic
- Other services with configurable Base URL + API key
See in-app AI provider settings for the live list.
Setup
- Settings → AI providers
- Add provider — name, base URL (if needed), API key
- Test connection
- Enable preferred models; set defaults (optional)
- Pick a model when creating tasks
Multi-provider strategy
| Scenario | Suggestion |
|---|---|
| Case design & long-doc analysis | Strong reasoning models |
| Formatting & list cleanup | Lightweight / low-cost models |
| Sensitive data | Private compatible gateway |
Keys stay on your machine—they are not sent to Ember’s release servers.
Security
- Keys live in local app data—protect the device
- Never expose keys in screenshots
- Use scoped sub-accounts and quotas for teams
FAQ
Connection issues: FAQ.