Connection errors can occur from time to time, even under the best circumstances.

First, simply try resubmitting your prompt. If your prompt is very large, it may continue to fail after multiple attempts. In that case, reduce the prompt size and try again.

Also, make sure you entered the model name correctly in CotranslatorAI. Go to Help > OpenAI > Models. On the OpenAI page that opens, find the model you want to use and compare its Model ID with the model name entered in CotranslatorAI. They must match exactly, including capitalization. Keep in mind that specialized image and audio models will not work in CotranslatorAI.

If the model name is correct, try a different model. A specific model may experience temporary problems even when other models are working. If other models work, double-check the name of the model that is not working and/or try using it again later.

If none of the models work, the OpenAI servers may be temporarily unavailable. Check the API status at https://status.openai.com/. If the servers are down, the only solution is to wait for them to come back online. However, OpenAI server issues are rare and usually do not last long. If your AI connection remains down for an extended period, OpenAI downtime is probably not the cause.

The issue could also be related to your setup, especially your Internet/wireless connection or router. Make sure your Internet connection is stable. If possible, switch to a different connection point and try again.

You should also check your OpenAI account balance and make sure it has sufficient funds. If it does not, add funds to your account.

Here is a reference guide for interpreting error messages generated by CotranslatorAI, along with troubleshooting options:

Invalid API key

  • What it means: The API key is missing, incorrect, expired, or revoked.

  • What you can do: Verify that the correct API key is configured under Tools > Options > OpenAI API.

Authentication failed

  • What it means: The request could not be authenticated.

  • What you can do: Verify that a valid API key is configured under Tools > Options > OpenAI API, and that the organization/project settings in your OpenAI account are correct.

Insufficient credits or quota exceeded

  • What it means: The account has run out of available credits or has exceeded its usage limit.

  • What you can do: Add billing or wait until quota is restored.

Rate limit exceeded

  • What it means: Too many requests were sent in a short period of time.

  • What you can do: Wait a few seconds and try again.

Model not found

  • What it means: The requested model does not exist or is unavailable to your OpenAI account.

  • What you can do: Select another supported model in the active AI Settings Profile and try again.

Invalid request

  • What it means: One or more request parameters are invalid.

  • What you can do: Check that all the settings in the active AI Settings Profile are compatible with the selected model. For example, unless the Reasoning Effort setting is set to (Auto-set), this parameter is only supported by reasoning models. Some models do not support a custom Temperature parameter. If you experience this error, try setting all the parameters in the AI Settings Profile to (Auto-set).

Context length exceeded

  • What it means: The prompt and expected response exceed the model’s maximum context size.

  • What you can do: Reduce the amount of input text or set the Max Tokens parameter of the AI Settings Profile to a smaller value unless it is already set to (Auto-set).

Content policy violation

  • What it means: The request or generated content violates OpenAI usage policies.

  • What you can do: Modify the prompt and try again.

Server temporarily unavailable

  • What it means: OpenAI’s servers are temporarily unavailable or overloaded.

  • What you can do: Wait and try again later.

Request timed out

  • What it means: The request took too long to complete.

  • What you can do: Try again later or reduce the amount of text being processed.

Network error

  • What it means: The application could not communicate with OpenAI due to a network problem.

  • What you can do: Check your Internet connection and try again.

Connection interrupted

  • What it means: The network connection was lost while the request was in progress.

  • What you can do: Try the operation again.

SSL/TLS error

  • What it means: A secure connection to OpenAI could not be established.

  • What you can do: Verify your network or proxy configuration.

Invalid response from server

  • What it means: The server returned data that could not be interpreted.

  • What you can do: Try again. If the problem persists, contact support.

Request cancelled

  • What it means: The operation was cancelled before completion.

  • What you can do: Run the request again if desired.

Unknown error

  • What it means: An unexpected error occurred.

  • What you can do: Try again. If the problem continues, contact support.

If the problem persists, please submit a support ticket, and we will try to guide you toward a solution.