Troubleshooting · Cursor

Cursor: Request Failed With Status Code 401 — Custom API Key Fix

A 401 inside Cursor chat means the endpoint answered and rejected the credentials: key and base URL mismatched, key revoked, or account access withdrawn. The fix path.

What you see

Request failed with status code 401

Why it happens

  • The key verified once but was later revoked or expired — Cursor keeps using it until the endpoint starts answering 401.
  • The base URL override was changed after the key was saved, so the saved key now goes to an endpoint that has never seen it.
  • The account behind the key was suspended or its access withdrawn.

How to fix it

  • Re-check the pair as it exists right now: the override URL and the key must come from the same issuer. Fix whichever side changed.
  • Test the exact key against the exact base URL with curl — it removes Cursor from the equation in one step.
  • If curl also returns 401, the key itself is dead: check its status in the issuer's dashboard or reissue it.

Reproduce outside Cursor

curl https://router.apitoken.sale/v1/chat/completions \
  -H "Authorization: Bearer sk-pool-•••" \
  -H "content-type: application/json" \
  -d '{"model":"gpt-6.1-terra","messages":[{"role":"user","content":"ping"}]}'

Related searches

  • {"error":{"message":"Invalid or missing API key.","type":"invalid_request_error","param":null,"code":"invalid_api_key"}}
  • cursor 401 custom api key

FAQ

Cursor verified my key earlier — why 401 now?+
Verification is a snapshot. A key revoked afterwards, or a base URL changed afterwards, produces 401s on every later request even though the initial verify passed.
Is this Cursor's bug?+
Almost never. A 401 is the endpoint's answer passed through. Reproduce it with curl: if curl gets 401 too, the credentials are the problem.
What if curl succeeds but Cursor still gets 401?+
Then Cursor is sending different credentials than you think — re-open settings and re-paste the key with the override URL already enabled.

Skip the broken setup

apiToken.sale serves GPT and Gemini models through OpenAI-compatible, Messages (Anthropic format) and native Gemini endpoints — standard SDKs, one key, one prepaid balance. Point your tool's base URL at it, and the config on this page works as written.