Connectors / OpenAI
AI
OpenAI
Add OpenAI's GPT models to your app with your own key, billed to your own OpenAI account.
Included on every plan. An admin approves it once.
What your app can do with it
- ✓Usage is billed by OpenAI to you, with nothing added by us
- ✓The key is kept by Syntica, never in the app's code or the browser
- ✓Set a monthly budget in OpenAI's console before you start
For your IT admin
Setting it up
Done once, in your company's own account with OpenAI.
1
Create a project for the application on the customer’s OpenAI account, so its key can be revoked on its own.
2
Create an API key in that project, with the least permission that works.
3
Set a monthly budget on the project. Usage lands on the customer’s bill, not the platform’s.
4
Hand the key to the platform once. It is never shown again.
For developers
What your app receives
Calls the OpenAI API with a key the customer owns. One variable, injected server-side at deploy time, so the key never reaches a repository or a browser, and never a client bundle, where a pasted key is public within the hour. Create the key in a project of its own so it can be revoked without touching anything else, and give it the least permission that works: a key that calls models does not need to manage the account. Usage is billed by OpenAI on the customer’s account, so set a monthly budget in the console before the first deploy.
OpenAI's own documentation →OPENAI_API_KEYGet your spot in the beta.
We open spots from the waitlist, oldest first. When yours opens, the free tier is a real environment with a real URL. No card.
We use cookies
We use essential cookies to run the site and optional ones to understand usage. See our Privacy Policy.