Film your hands and your pan, then write the recipe the way you would say it. Script mode reads it in a steady voice, waits exactly as long as you tell it to while the footage plays, rings a bell when the step changes and gives you one file per step. Change an amount later and only that sentence is made again. Want one voice or two hosts instead? The same Studio does that too.
Abd Shanti's one pan shakshuka, read by one voice, by two hosts, and as timed steps with a bell and a pause for the footage. The line being read lights up; click any line to jump to it.
The whole recipe in one voice, the way most short recipe videos are narrated. One paragraph in, one MP3 out, with the same half second between every sentence.
Drawn to show where the voice lands on a Short. The sound is the real recording.
(pause 3) anywhere in a Script mode script for exactly 3 seconds, up to 60. Sliders set the pause after every comma, sentence and paragraph.(ding) 0.9 s, (ding 0.5) shorter, (beep) 0.25 s, (beep 2) up to 10 s, (click) 0.06 s. A sound plays right after the words before it.TRACK: Step 2 on its own line starts a new file. Download one file per track, every track as MP3s in a ZIP, or one file with a 2 second gap between tracks.Hands only recipe videos are often shot on a phone from above, with the voice added after. Script mode works in that order: the footage exists, and the voice is written to fit it.
“Decide early whether you'll speak on camera, record a voiceover later, or rely on text overlays to guide viewers.”
One step per line, with amounts in words: two tablespoons, not 2 tbsp. Name a step in plain words, like Step one, and never with a capitalised label before a colon.
Write (pause 3) where the footage needs time, (ding) where a step ends, and TRACK: before a step when you want it as a file of its own.
Choose one of the 22 steady voices. If it says a dish name wrong, add the word under Say it right with how it should sound.
Press Make recording, download the MP3 or WAV files, and drop each one on its clip in your video editor.
(pause) after the step.INGREDIENTS: or STEP 1: is read as a character name, and the recipe turns into a scene with a voice per name. Write Here is what you need. or TRACK: Ingredients.CUE: Step two. on its own line, or [Step two] inside a line, is read in the cue voice.Cooking takes the time it takes. The voice should wait for the pan.
Recipe edits are fast: one step is a two second clip, the next a ten second one. A guide to writing cooking scripts puts it plainly: “Nothing loses viewers faster than a rushed or slow video. Your script should estimate how long each segment will take” (How to write a script for a cooking video).
Done by hand, that means recording each step, cutting the takes apart and sliding them along the timeline until each one lines up with its clip.
Script mode writes the timing into the script. Every (pause 3) is three seconds of silence exactly where you wrote it, a bell marks each change of step, and every TRACK: becomes a file of its own that you drop on its clip.
Bars show the longest time in each range. Put TRACK: before each part and Script mode shows every part's length before you record, so you can hold your script up against a plan like this one.
TRACK: IngredientsStarts a file of its own. Not read out.Two tablespoons of olive oil, (pause 0.5) one onion, (pause 0.5) one red pepperRight after a comma, the pause replaces the comma pause, so each ingredient gets exactly half a second.TRACK: Step 1, the baseA new file for this step's clip.(ding) Warm the olive oil over a medium heat. (pause 1)At the start of a track the bell opens the file. Then 1 second before the next line.Add the onion and the pepper, and let them soften for five minutes.The last line of a track. The file ends here, with no pause after it.TRACK: Step 4, the eggsStep four, its own file again.(ding) Make four wells with a spoon, and crack an egg into each one. (pause 2)Two seconds for the eggs going in, then the next line.(pause N), from 0 to 60 secondsExactly where you write it: after a step, after a comma, even between two words.
(ding) plays right after the step's last words, then the pause. At the start of a track it opens the file.
TRACK:Each step or each Short is its own file, so you never search through one long recording in the editor.
Each track shows its estimated length, so a step that runs long gets trimmed before anything is made.
(pause) and press Update recording. Nothing is recorded again.Adding a sound after a line, or moving it, never records your lines again.
(ding)Bell 0.9 sThe step is done. Write (ding 0.5) for a shorter bell.(beep)Short beep 0.25 sNow: add the eggs, flip it, take it off the heat.(beep 2)Long beep Up to 10 sA timer going off, or the end of a part.(click)Click 0.06 sA small action, like a pinch of salt or a lid going on.A sound plays 0.15 seconds after the words before it, and a pause you wrote comes after the sound. At the very start of a track, a sound opens the file. In the Sentences view of Script mode, click the pause after any line to add a bell, beep or click there, or to take one out.
Plain text, no special editor. Paste from your notes app or your blog and it keeps working.
TRACK: Intro Shakshuka for two. (pause 1) One pan, twenty minutes, and eggs that stay soft. TRACK: Ingredients Here is what you need. Two tablespoons of olive oil, (pause 0.5) one onion, (pause 0.5) one red pepper (pause 0.5) and two cloves of garlic. A teaspoon of cumin (pause 0.5) and a teaspoon of paprika. One can of chopped tomatoes, (pause 0.5) four eggs, (pause 0.5) salt (pause 0.5) and a handful of parsley. TRACK: Step 1, the base (ding) Warm the olive oil over a medium heat. (pause 1) Add the onion and the pepper, and let them soften for five minutes. TRACK: Step 2, the spices (ding) Stir in the garlic, the cumin and the paprika. (pause 2) One minute is enough. TRACK: Step 3, the sauce (ding) Pour in the tomatoes, add a pinch of salt, and let it simmer for ten minutes. TRACK: Step 4, the eggs (ding) Make four wells with a spoon, and crack an egg into each one. (pause 2) Cover the pan, and cook until the whites are set and the yolks still move. TRACK: Serve (ding) Parsley on top, (pause 0.5) warm bread on the side, (pause 0.5) and eat it straight from the pan.
Seven tracks, one per part. Say it right: shakshuka as shakshooka. Script mode estimates the whole thing at about a minute and a half.
| You write | What happens |
|---|---|
TRACK: Step 2, the spices | Starts a new audio file with that name. The line is not read out. |
Step two. | Names a step in plain words. It is read out like any line. |
(pause 3) | Exactly 3 seconds of silence. (pause 0.5) works too, up to 60 seconds. |
one onion, (pause 0.5) one red pepper | Right after a comma, the pause replaces the comma pause. |
(ding) (ding 0.5) | A bell, 0.9 seconds or as long as you write, from 0.2 to 5. |
(beep) (beep 2) (click) | A sound, right after the words before it. |
CUE: Step two. or [Step two] | Read in the cue voice, a second voice you choose. |
Two tablespoons of olive oil | Amounts written the way you say them, so the voice never has to guess. |
Tip: stir it often. | Tip, Note and Warning before a colon stay part of the line and are read out. |
INGREDIENTS: or STEP 1: at the start of a line | Avoid it. A word in capitals before a colon becomes a character, and the recipe turns into a scene. |
~~~~~ | A line with no words is decoration. It is skipped. |
Say it right: shakshuka as shakshooka | Said your way everywhere in the script. The script stays as written. |
You cook it again, and the sauce needs longer. The voiceover should not have to start over.
let it simmer for ten twelve minutes
1 sentence · 80 characters to record · nothing else changes
Edit the text and press Update recording. Only the sentences you changed are recorded again, and every other sentence keeps the exact audio it had. The status line tells you what will be recorded before you press anything.
The pauses and sounds after your lines, and the order of the lines, are not part of any recording. Make a step's pause longer, add a bell after it or move a line, and the file is rebuilt from the audio you already have.
It works because Script mode only uses steady voices: a steady voice says the same sentence the same way every time, so the new amount sits in the track like it was always there.
No brand names and no prices, just how each way works in practice.
| What you need | Made for timingFreeTTS Script modeFreeTTSPaste the recipe, press record | Recording your own voiceYour voiceA phone mic, after filming | Hiring a voice actorVoice actorA booked session | A general text to speech toolOther TTSPaste text, get one voice |
|---|---|---|---|---|
| Ready any timeNo quiet room, nobody to book | Yes. Any time, in the browser | Partly. Needs a quiet kitchen | No. Book it and wait | Yes. Yes |
| Pauses sized to the footageTime for each step | Yes. Written in the script | Partly. Cut by hand in the editor | Yes. They read to your cut | Partly. Some offer pause tags |
| A sound at each stepA bell when the step changes | Yes. (ding) in the script | Partly. Added in the editor | Partly. Added in the editor | Partly. Added in the editor |
| One file per step or ShortDrop each on its clip | Yes. TRACK: per step | Partly. Cut it yourself | Partly. If you ask for it | Partly. Split the text by hand |
| Dish names said rightShakshuka, za'atar, gochujang | Yes. Say it right | Yes. You know them | Yes. Give them notes | Partly. Spell it out and hope |
| Change one amount laterTen minutes becomes twelve | Yes. Only that sentence | Partly. Record again, match the sound | Partly. A pickup session | Partly. Usually make it all again |
| Sounds like youYour own personality on the channel | No. A synthetic voice | Yes. It is you | Partly. A real person, not you | No. A synthetic voice |
| Open Script mode |
Film from above on your phone, then voice it in the evening, with no need for a quiet kitchen.
One TRACK: per Short, each checked for length before you record. YouTube counts vertical videos up to three minutes as Shorts.
Recipe card plugins can hold a video in the card and its schema. Voice it from the card's own steps.
A (beep 2) where the timer would go off, and pauses that match the cut, not the hour the dough took.
A steady voice reads it the same way each time, and changing one amount records one sentence.
Dialogue gives each host a voice of their own. A free account can try one short scene a day.
The Script mode settings are in the panel on the right. Pauses and sounds after a line never record anything again. Voices and speeds do, because they change the sound.
(ding 0.5) (pause 1)(pause 3)Abd: line, one per lineOn September 29, 2026 we read the code behind each feature and ran the scripts on this page through Script mode's own parser.
Pauses, sounds, tracks, Say it right and Update recording from Script mode's parser and its controls; speeds, formats and limits from the Studio and the voiceover tool; voice and language counts and prices from the site's own list of facts.
Script mode reads it as seven tracks with a bell opening each step, no line mistaken for a character, and an estimated length of 1:32.
We ran INGREDIENTS: two eggs, STEP 1: crack them. and Now whisk. through the same parser. Both labels became characters with voices of their own, and Now whisk. became a stage direction in the cue voice.
Changing ten to twelve in the sauce step leaves one changed sentence of 80 characters, which is what the status line counts.
All three were recorded with Script mode on September 29, 2026, with the voices and settings listed under each, and nothing was edited afterwards. In the timed steps file every bell runs 0.9 seconds right after its step, and the silence after it matches the written pause to within 0.05 seconds. More on how exact the pauses are on the actors page.
Backstage: How to make cooking videos. Decide early between speaking on camera, a voiceover recorded later or text overlays; overhead shots for chopping and stirring.
Vox Ghostwriting: How to write a script for a cooking video. Estimate how long each segment will take, with a timing guide from the intro to the end message.
The Daily Gazette: How BuzzFeed's Tasty conquered online food. Overhead videos where “a disembodied pair of hands moving at bullet speed” makes the dish.
YouTube Help: YouTube Shorts length. Since October 15, 2024, square and vertical videos up to three minutes count as Shorts.
Pinterest Newsroom: Introducing an easier way to create. Since May 10, 2023, video and image posts are all one format, the Pin.
Mediavine Help: Video player best practices, on embedding the video in the recipe card. WP Recipe Maker: Recipe video, which adds it to the recipe schema.
YouTube Help: Channel monetization policies, including the inauthentic content policy (renamed July 15, 2025), and Disclosing altered or synthetic content.
FreeTTS, September 29, 2026: the Studio's code and Script mode's parser, as described under How we checked.
FreeTTS Script mode turns a written recipe into a cooking voiceover: exact pauses for the footage, a bell at each change of step, one audio file per step and dish names said the way the creator writes them. Changing one amount records only that sentence again.
Paste the shakshuka script, or your own recipe, into Script mode and see how long each step runs. Pasting and timing is free; recording is part of PRO.