Upload a video — write each line with its exact time, or let AI find where narration belongs, write & correct the script, and add the voice at the right moments. Free, no signup, no watermark.
Drag & drop your video here or browse files
MP4, WebM, MOV — best under 5 minutes / 200 MB (browser memory)
Plenty of good videos never get watched because nobody wants to record their own voice — the room is noisy, the microphone is poor, or it just feels awkward. This tool gives your video a voice anyway. Upload the video, decide where the narration goes, let the AI tidy up the words, and download the finished file with the voice already mixed in. No recording, no studio, no software to install.
There are three ways to place the voice, and the first is the one most people use. Timed Script is for anyone who already knows what they want to say: you write the lines, each one starting with the time it should be spoken, and every voice lands at exactly that second. Retake is for videos that already have speech — it listens, writes down what was said, cleans up the wording, and re-voices those same moments in a fresh voice. Narrate is for silent footage and B-roll: it finds the scene changes and the quiet stretches, and the AI writes one short line per scene around a topic you give it.
Step 1: Upload your MP4, WebM or MOV file. Shorter is smoother — under about five minutes and 200 MB works best, because everything runs in your browser.
Step 2: Pick a mode. Timed Script is fastest if you have your lines ready; Retake and Narrate detect the moments for you.
Step 3: Review the script. The AI can write it or just fix the grammar, but read every line before you go on — the AI fixes language, not facts.
Step 4: Choose one of the 16 free download voices, decide what happens to the original audio (replace it, duck it quietly underneath, or keep it), and generate the voice track.
Step 5: Download the finished MP4. You can also take the narration on its own as a WAV file, or the timed script as an .srt file.
Tutorials and demos: record the screen silently, then add calm, even narration line by line with Timed Script. Product and business videos: give a silent product clip a voice that explains the offer. Travel and B-roll: let Narrate mode watch for the scene changes and write around your topic. Fixing a rough recording: if the original narration is mumbled or full of mistakes, Retake mode re-voices it cleanly in the same places. Faceless channels: a steady, watermark-free workflow for videos where nobody wants to speak on camera.
The listening, the scene detection, the voice generation and the mixing all happen on your own device; your video file is never uploaded to our server. The one exception is the optional AI script help: when you press an AI button, the script text is sent to the AI service so it can write or tidy your lines. The tool tells you this on the page before the button, not after — and Timed Script mode works from start to finish with no AI involved at all.
Yes, completely. There is no signup, no credit card, no daily limit and no watermark on your download. The voices are free open-source voices and the editing runs in your own browser, which is how we can keep it free.
You choose. In Timed Script mode you write each line with its start time and the voice lands at exactly that second. In Retake mode the tool transcribes the speech already in your video and re-voices those moments. In Narrate mode it finds the scene changes and quiet stretches in silent footage and writes one line per scene.
Yes. That is Timed Script mode, the fastest way to use this tool. Write one line per moment, starting with its time - 0:05, [0:12], 1:20, 00:01:20, 5s and a plain 90 for seconds all work - and each line's voice starts at exactly that time. If a line is too long for the gap before the next one, the tool warns you and speeds the voice up a little instead of cutting it short.
You pick from 16 free downloadable voices - American, British, Scottish and other accents, plus female, deep, elderly, robot and whisper styles. They are built on the free eSpeak engine, so they sound clear but synthetic, not like a human actor or a paid studio AI voice. That honesty is why they are free with no API key and no limits. The natural voices from your own device can be used to preview lines, but browsers do not let any website record them, so they cannot be baked into the download.
The analysis - listening to your video, finding scenes, placing the voice - happens on your own device. Your video file itself is never uploaded. Only if you press an AI script button is the script text sent to the AI service so it can write or tidy your lines, and the tool tells you this before the button, not after. Timed Script mode works fully without any AI at all.
English is fully supported. Urdu script-writing works too, but we show a warning before you start because its accuracy can vary, and there is no Urdu export voice yet, so the downloaded narration is English for now. For the smoothest result use videos under about five minutes and 200 MB - everything runs in your browser tab, so very long or 4K files can run out of memory. MP4 videos give the most reliable finished download.