Turn your voiceover into a timed transcript.

Upload the narration you generated in Google AI Studio. Whisper runs locally in your browser and creates an editable, timestamped transcript ready for ChatGPT and Course Slide Studio. No account, paid API, or uploaded audio.

01 / Audio & settings

↥Choose or drop your voiceoverMP3, WAV, M4A, OGG and browser-supported audio

No file selected

Recognition quality

Models download on first use and are usually cached by your browser. Larger models can be slow on older devices.

Timestamp detail

Word timestamps are estimates. Always review key slide transitions against the audio.

Ready. Choose an audio file to begin.

Speech recognition runs on your computer. The first model download requires an internet connection.

Transcript & timings

Click a timestamp to listen, then edit any misheard words.

No transcript yet

▥

Your transcript will appear hereWord timings, corrections and downloads in one place.

Student workflow: 1. Create your script in ChatGPT. 2. Generate narration with Google AI Studio. 3. Transcribe that exact finished audio here. 4. Download the timed SRT and attach it to the GPT lesson-slide prompt. 5. Import the resulting JSON into Course Slide Studio.

Limitations: This tool needs a modern browser and enough device memory. Word timings can drift near pauses or long recordings; check your finished video against the narration. Nothing is charged per transcription, but downloads use normal network bandwidth.