Write a bedtime story for [child's name], age [age], who loves [3 things, e.g. dinosaurs, fire trucks, her dog Biscuit]. [Name] is the hero. Tonight's theme: [e.g. being brave on the first day of school / sharing / trying new food].
Rules:
- [5] minutes read aloud (about [700] words), calm and getting sleepier toward the end
- simple words a [age]-year-old understands, one gentle repeated phrase they can say along
- include [a friend/sibling/pet name] as a sidekick
- no scary villains; the problem is solved with kindness or cleverness
- end with [Name] falling asleep somewhere cozy
Format it for narration: short paragraphs, no headings, and put a [pause] where the narrator should slow down.
Steps & tips
Paste this prompt into any chatbot (ChatGPT, Claude) to write the story, then paste the story into ElevenLabs Text to Speech or Studio.
Pick a soft, slow voice and lower the speed a little; play it back before bedtime to check.
Never clone someone's voice without their permission.
ElevenLabs
Translate a video into another language while keeping your voice and timing.
Use cases you didn't know about
Dub your video into Spanish in your own voice
Who it's for
Creators who want to reach viewers in other languages
(Upload your video to ElevenLabs Dubbing, then use this as your checklist and review notes.) Source language: [English]. Target languages: [Spanish, Portuguese, Hindi]. Speakers: [number]. Before exporting, check:
1. Names, brand names and slang I want left untranslated: [list].
2. Terms that need a specific translation: [term = translation].
3. Lines where the joke or idiom won't translate: rewrite them with a local equivalent.
4. Listen to the first 60 seconds and the call to action in each language with a native speaker if possible.
After export, write a translated title, description and 3 pinned-comment ideas for each language.
Steps & tips
ElevenLabs Dubbing keeps speaker voices and supports dozens of languages; editing the transcript fixes mistranslations.
Upload dubbed versions as extra audio tracks on YouTube if your channel has multi-language audio.
Only dub voices you have the rights to.
Descript
Correct a wrong word in your recording by editing the transcript, with AI speech in your own voice.
Use cases you didn't know about
Fix a flubbed podcast line by typing it
Who it's for
Podcasters and YouTubers who misspoke and can't re-record
(In Descript, edit the transcript and use its AI voice to regenerate the corrected words.) Before you fix it, paste your transcript section into any chatbot with this: "Here's a transcript excerpt from my [podcast/video]: [paste]. I misspoke at: [quote]. Rewrite only that sentence so it's correct and sounds like natural speech in my style (casual, [short sentences]). Keep it within [number] words so it fits the gap, and give me 2 alternatives." Then paste the best version into Descript.
Steps & tips
Train Descript's AI voice only on your own voice, with your consent; it can't be done for someone else.
Keep fixes short (a few words) so they blend in, and listen at full volume before publishing.
Use Studio Sound afterwards to even out the audio.
ElevenLabs
Design distinct character voices and generate audio lines to play at the table.
Use cases you didn't know about
Give your D&D NPCs real voices
Who it's for
Game masters who want their villain to actually sound scary
(Use ElevenLabs Voice Design: describe the voice, then generate lines in Text to Speech.) Voice description: [a gravelly old dwarf blacksmith, slow and warm, slight Scottish lilt / a cold, whispering elven queen, precise and menacing]. Lines to generate:
1. Introduction: "[NPC greeting]"
2. Key information: "[clue or quest hook]"
3. Reaction if the players are rude: "[line]"
4. A dramatic reveal: "[line]"
Add emotion cues in brackets where supported, like [whispers] or [laughs].
Steps & tips
Generate a few voice options and save the one you like for the whole campaign.
Pre-make lines before the session so you're not fumbling at the table.
Keep improvised dialogue for your own voice; the audio works best for big moments.
Otter.ai
Turn a recording or a long PDF into the key points and action items.
Summarize this meeting for people who weren't there: (1) 3-sentence summary, (2) decisions made, (3) action items as a table with owner and due date, (4) open questions. Keep it under 250 words and quote anyone verbatim only where wording matters.
Steps & tips
Connect Otter to your calendar so it joins and records automatically.
Tell attendees you're recording; follow local consent rules.
Paste the prompt into Otter's AI chat after the meeting.
source
40% of workers who've used AI chatbots at work have used them to summarize documents or meetings.
[Your script here]
(Voice settings to try: a warm, conversational narrator voice, stability around 50%, then adjust. Add commas and ... for natural pauses, and spell out numbers the way you want them read.)
Steps & tips
Paste the script into Text to Speech and choose a voice.
Generate a few takes and pick the best one.
Only clone a voice you own or have permission to use.
source
ElevenLabs ranks #25 in a16z's top 50 consumer AI web products and #10 in consumer spending.