Skip to main content
API keys authenticate model requests. Create separate keys for different applications or purposes so you can track usage and disable access individually.

​
Create a key

  1. Open API keys in the console and confirm the account and workspace.
  2. Create a key with a recognizable name.
  3. Set the model scope, allowance, and expiration as needed, then save.
  4. Copy the key into an environment variable or secret manager and follow the quickstart.
If the creation dialog says the full key is shown only once, store it before closing. Keep full keys out of frontend code, public repositories, and support tickets.

​
Advanced settings

After saving, verify the configuration with a real request. Settings do not add input, output, or tool capabilities that the target model lacks.

​
Disable, replace, or delete

  • Disable: Temporarily stop requests with that key. Check validity and expiration before re-enabling it.
  • Replace: Create a new key, update your application, verify requests, and then disable the old one.
  • Delete: Confirm that affected applications have switched credentials before completing deletion.
Enterprise workspaces offer member and system keys. See that guide for ownership, permissions, and expiration rules.

​
If requests are restricted

Check the key’s status, expiration, model scope, remaining allowance, and account balance. Enterprise requests also follow workspace budgets. Review the error and Tid in request logs instead of repeatedly retrying an invalid credential. To manage account resources, use a management key. Last updated: 2026-10-13