When a model you added with your own API key returns an error, the service turned the request down: check the API key, the model, the endpoint and your quota with the provider, then try again.
API unavailable beside a model in Settings › Text Model or Transcription Model means the model has no API for that kind of model, for example after you deleted its API in Settings › API. Swipe the model, tap Edit, and choose another API, or tap Manage APIs to add one (add your API key).
What to check
- An HTTP error, such as "The text service returned HTTP 401": the key may be wrong or expired, or the model, endpoint or quota may not match your account. A Google Cloud access token expires and must be replaced.
- No complete answer from a text model: choose a chat model that supports JSON responses.
- A recording too large or too long for the provider: transcribe it with another model, such as Apple Speech, with Regenerate… on the recording (choose the Transcription Model).
- A retired model: AideMemo names the model to choose instead. Edit the model in Settings › Transcription Model and choose it.
- No model list in Add Model: check the key and its permissions, then Retry Loading Models, or use Enter Model ID.
AideMemo AI's messages
- Out of credits, or today's limit reached: Get More Credits, or wait for the refill or the next day (out of credits).
- Too many requests, or busy: try again in a minute or a few minutes.
- Couldn't reach AideMemo AI: check your internet connection and try again.
If an error keeps coming back, send it to us with Report App Issue.
This page answers
- Why does my API model return an HTTP error?
- What does "API unavailable" mean?
- My own transcription service fails
- AideMemo AI says it's busy or out of credits
Something else? See all help articles, or contact us.