Models and tools
OpenAI
An OpenAI integration stores a project-owned OpenAI API key so Mobius-hosted agents can run GPT models through that key.
Use OpenAI when a project should use its own OpenAI account and quota for agent model calls. The integration does not add source events.
Capability map
| Capability | Value |
|---|---|
| Provider ID | openai |
| Auth kind | api_key |
| Connect flow | Inline API key |
| Actions | No |
| Events | No |
| Webhook delivery | No |
| Event samples | No |
| Live status | No |
Connect OpenAI
| Surface | Support |
|---|---|
| App | Open Govern > Integrations, choose OpenAI, and paste an OpenAI API key. |
| CLI | The mobius CLI does not connect OpenAI integrations yet. Use the app or API. |
| API | Call POST /v1/projects/{project}/integrations/providers/openai/connect with api_key. |
A project-owned key takes precedence over the platform credential for OpenAI models. This lets paid projects keep agents running through bring-your-own-key credentials even when platform-funded model calls are capped.
Actions
OpenAI does not register action catalog entries. Use it through agent model selection.
Events
OpenAI does not register source events.
Next
- Configure model use from agents.
- Review model routing in steps.
- Keep provider keys in integrations.