Choose a file
Upload an audio file, record a voice memo, or extract a supported video’s audio locally before transcription.
Upload an audio file, record with your microphone, or extract the audio track from a compatible video in your browser. Transcribe, edit, summarize, or create manually timed subtitles.
Clean and Summary each use one of your 2 daily AI text actions. Your selected text is sent to OpenAI. AI output may omit details; always compare it with the original.
Create timed subtitles from your transcript. Play the audio and mark start/end times yourself; automatic timestamps and speaker detection are not available.
Timing and speaker labels are entered manually. Click Play on a caption to seek to its marked start time.
Review names and numbers for accuracy.
Upload, review, and export. You stay in control of the final transcript.
Upload an audio file, record a voice memo, or extract a supported video’s audio locally before transcription.
Choose Auto-detect, English, Mandarin or English + Chinese. Check the transcript alongside the original recording.
Export TXT, DOCX or Markdown, or manually mark subtitle timing and download SRT or VTT.
Upload audio, record a WAV voice memo, or extract audio from a compatible video locally in your browser.
Convert spoken content into text using an AI transcription model.
Choose automatic detection, English, Mandarin, or bilingual English–Chinese hints.
Provide expected names, acronyms, or specialized terms to guide recognition.
Edit the transcript, generate an optional AI-cleaned version, and compare both without replacing the original.
Create optional AI notes, export text, and manually time SRT/VTT captions without invented timestamps.
There is a limited free allowance. Availability depends on current daily and overall service limits, which protect the shared transcription budget.
Upload completed MP3, WAV, M4A, FLAC, OGG, MP4 or supported WebM audio. For compatible MP4, MOV, MKV and WebM videos, use on-device audio extraction first. Browser recording creates a WAV; codec support varies.
Each file can contain up to 15 minutes of verified audio and must be no larger than 24 MB.
Yes. Choose Mandarin, English, or English + Chinese when those languages are expected. Automatic detection is also available. Always check names, numbers and code-switched speech.
Yes. Edit the original and AI-cleaned versions separately, create optional summaries, or manually time subtitle cues. Download TXT, DOCX, Markdown, SRT or VTT. AI text operations have a separate daily limit.
Audio is uploaded and sent to OpenAI when you choose Transcribe. Optional AI cleanup or summary sends your selected transcript text to OpenAI separately. The site handles uploads as temporary processing files; editable results remain in the current tab unless you download them. Avoid submitting recordings you are not authorized to share.