Skip to main content

Voice input

2.6K
When making voice recordings, it is often customary to say verbal and intonational words, which can then be converted into written language using ChatGPT to optimise the speech-to-text effect. Additionally, it can also be used to organize disordered text.
Prompt Content
Using concise and clear language, please edit the following passage to improve its logical flow, eliminate any typographical errors. Be sure to maintain the original meaning of the text. Please begin by editing the following text: [voice input transcript]
255 chars · ≈ 75 tokens
FAQ
When cleaning up a voice-to-text transcript, will it strip my verbal tics too?
Yes. The AI defaults to removing filler words like "um," "you know," and "I mean." If you want to keep certain tics as part of your personal voice, say explicitly "keep the tic X as part of my speech rhythm, do not remove it," otherwise the output will feel overly written.
Can I clean up a transcript of several thousand words in one go?
Not recommended. GPT tends to cut corners or even skip entire middle sections on very long input. Break it into chunks by topic or every 2000 words, clean each piece, then do a final pass over the whole thing for transitions.
How do I use this prompt?
Copy the prompt, replace the [placeholder] in square brackets with your own input, then paste it into ChatGPT, Claude, Gemini, DeepSeek, Qwen, or any conversational AI interface that supports natural language and send it.