AI voices sound human now. But they still guess names from the spelling, and nobody has time to listen to every line. Onepin runs on the voices you already use (ElevenLabs, OpenAI, Google, 30+ more), spells out prices and dates, checks names against a 4M-word pronunciation dictionary, scores every line, and fixes the wrong word without re-rendering the take. Built for teams shipping AI voiceover in several languages: product videos, dubbing, courses. Free to start, no credit card required.
Free Options
Launch Team / Built With
AWS Startup AdvisorPersonalized AWS guidance that meets you where you’re already building
Promoted
Hey Product Hunt 👋 SB here, founder of Onepin. Thanks @rohanrecommends for hunting us.
A great AI voice reads "Porsche Taycan" as TAY-can. Porsche says TIE-kahn. It guessed from the spelling, and nobody caught it, because nobody listens to line 1,200.
We spent two years at Podonos measuring how AI voices sound, with 150,000 human listeners doing the rating. The models solved delivery. What kept failing was names, prices and dates, and the only check most teams run is a person with headphones.
Today we're launching Onepin, the production step after text-to-speech. It runs on the voices you already use, ElevenLabs, OpenAI, Google and 30+ more, and checks every line before it ships.
Here is what it does to every script:
🔢 Spells out prices, dates and abbreviations before the voice reads them.
📖 Checks names, brands and drug terms against a 4-million-word pronunciation dictionary.
📊 Scores every line for naturalness, clarity, word accuracy and pronunciation, against a bar you set.
🔧 Fixes the one wrong word in the same voice, without re-rendering the take.
Two decisions we made on purpose. Onepin checks how a name was said, not just what the transcript spells. And there is no box where you type phonetic spellings: the dictionary and the fix do that work, so your script stays as written.
Built for anyone shipping AI voiceover at volume: product videos, ads, explainers, dubbing, audiobooks, courses, and platforms that put a voice in front of millions of users. API, SDK and MCP from day one. SOC 2 Type II.
You keep your voice. Onepin makes sure it says the right words.
Free to start, no card: onepin.ai. You can run one line on the homepage without an account.
Try it and tell me what's missing, here or at sb@onepin.ai. We'll be in the comments all day, answering questions and running your lines. 🚀
🎁 LAUNCH SPECIAL: 20% off for Product Hunt users through October 31. The code is under the 20% off badge above.
One question for you: what's the word your AI voice keeps getting wrong? Run it on onepin.ai (free, no card), switch the run to shareable, and paste the link below so everyone can hear the before and after. I'll reply to every one. No time to sign up? Drop the word and I'll run a few today.
@soohyun Many congratulations Soohyun on the Onepin launch! 😊
How I met the maker: I met the maker through a Product Hunt friend, who asked whether I’d be interested in checking out and hunting the product. The moment I saw Onepin’s landing page, I knew I wanted to hunt it, it was addressing a very real and often overlooked pain point.
What is Onepin: Onepin is a production layer for AI text-to-speech. It works with voices you already use, including ElevenLabs, OpenAI, Google, and 30+ others, and reviews every line before you ship it.
It spells out prices, dates, and abbreviations correctly, checks names, brands, and specialized terms, scores each line, and fixes mistakes in the same voice without leaving the page or re-rendering the entire take. That combination is what makes it unique.
Why I endorse it: AI voices can sound great, but one mispronounced name, price, or date can undermine an otherwise polished video, course, ad, or voice agent. Onepin turns that manual, error-prone review process into a scalable quality check.
The product is clearly already getting attention, one friend even told me I should hunt it after seeing it on social media, not realizing I already had. 😄
@rohanrecommends Awesome. Thank you for hunting Onepin. It is a real problem in the industry, we are solving this in a practical way, saves 80% of voice AI cost in production.
@marissa__bronfman Marissa, people typically pay 2x-9x more for one script because of regenerations. Onepin makes TTS to generate correctly or fixes them. Try. Directly contact me for any questions or suggestions.
@soohyun @rohanrecommends Fixing one of the wrong without redoing the whole recording is a really useful. Does the corrected words sound natural with the origina; voices???
@rohanrecommends @vipul_kumar1280 Yes, it is natural. Check this https://go.onepin.ai/hear-it
Our AI model instantly clones the voice & style and speaks correctly.
@rohanrecommends @soohyun I tried it out, as was looking for voice models for my product. Can I preview the voice by entering my script on the platform itself? or it's available for APIs only as of now?
second, does it detect the expressions like [breath],[pause], and so on.
Congrats for the launch!
@rohanrecommends @aryan_thakore Under Voice tab, you can type "best voice for <YOUR PRODUCT>", or ask Quincy "Find the best voice for <YOUR PRODUCT> here is my script: ...", you can see the relevant voices. Once you find the voice, build a template of the workflow, and you can reuse the workflow via API.
Currently it doesn't detect paralinguistics yet, it is under our roadmap. We have been focusing on pronunciations. Paralinguistics are on our roadmap.
@soohyun Great to see a tool that focuses on accuracy, not just voice quality. Does it support multiple languages?
@rohit62661 You are right, many of them are natural with minimal noise. Onepin supports 10 locales now, will expand to 30. Which languages are critical for your use case?
porsche as TAY-can lol 😂 finally someone fixing the names and prices thing. works with elevenlabs voices too?
@petrkovacik Yes, Onepin supports 30+ models including ElevenLabs. If you already subscribe ElevenLabs, bring the API key, then you can use your cloned voices in Onepin.
@rohanrecommends @hamza_afzal_butt Our agent always finds the right pronunciation and use it. Yet, there are words with multiple pronunciations, you can direct them how to pronounce just in case.
Congrats, Soohyun. I ship short product videos with AI voiceover and the pronunciation check is the step I do by ear today, so I will try this. Thanks a lot.
Any suggestions if I am already using ElevenLabs?
@nicklaunches Yes it is a real pain. We already work with a video companies, produced TikTok and YouTube videos using Onepin. If you have any questions or suggestions, please feel free to contact me directly.
Hi, great product. Planning to support Turkish language?
@aytekin_meral Yes. Turkish is one of the languages we will support next. What is your use case?
I have experienced firsthand the exact situation that you have described, a video for a class that had the name of one of the students mispronounced because no one bothered to go through the important words list to check if it contained names not listed there. Scoring each line of dialogue is the answer, not a quicker way to do the manual checking.
@hamza_aafzal you are right. Speaking generally well is not enough. Speaking names correctly (+numbers, currency, DHL tracking numbers) are the key. Try and let me know directly for any questions or suggestions.