#8447 You selected LLM error correction and used the OpenAI ChatGPT channel, but did not fill in the key or api url. Please fi

This is caused by turning on LLM Correction while the LLM correction channel was set to OpenAI ChatGPT, but no key or API URL was configured for it. The software therefore cannot call the model to fix the recognized text.

Choose one of the following:

Option 1: Turn off LLM Correction (fastest)

1. On the main window, in the speech recognition row, clear the LLM Correction checkbox.
2. Click Start Execution again.

Option 2: Keep LLM Correction and fill in a usable key (recommended)

1. Open the top menu Translation Settings → ChatGPT (or OpenAI ChatGPT), enter a valid API Key. If you use a relay API, also enter the API URL; otherwise leave it blank.
2. Save and start again.

Option 3: Switch to another configured AI channel

1. Open Menu → Tools/Options → Advanced Options → Speech Recognition (ASR).
2. Change LLM Provider for Proofreading (LLM Correction Channel) to a channel for which you have already filled in a key, for example DeepSeek.
3. Go to Translation Settings → DeepSeek, fill in the Secret Key, save, and start again.

Note: The LLM correction prompt is located at software directory/videotrans/prompts/resegment/llm.txt; this error has nothing to do with the prompt itself, only with the channel credential.

Please refer to the related documentation:

• Advanced Options
• Editing prompts
• ChatGPT translation
• DeepSeek