Spoken AI is a Large Language Model (LLM) powered Language Translator leveraging advanced architecture and comprehensive databases to deliver native-level translations in over 140 Languages and innovatively introducing Dialect-specific Translations
Bonjour Hunters 😉
Freelance Translators have often spoken of how machine translators have become way more accurate over the years, leading to a lower demand for their services. With the surge of machine learning, the inevitable was clear.
I'm here to say they were right.
Introducing Spoken AI 🌐文
🌍 Built on the Largest Language Model, supporting over 140 Languages & 130 Dialects.
🎯 True AI-Generative Translations: instant translations? not quite yet, each translation takes time to ensure accuracy and a natural read.
🗣️ World's first full-scale Dialect Support: Translate into Mexico's Spanish or Shanghai's Chinese and much more.
Introducing Accents!
assimilate text into the desired accent (e.g English to Shakespearian English ➡️ "Good morrow, how art thou this day?".
OUR STORY!! 📖
I started Spoken AI to pioneer Dialect-Specific Translations. But what I've found proved to be far greater than first intended.
In creating our LLM, we realized our translations could not be differentiated from native speech, even at the vernacular level. we trained our LLM in a vast library of languages and incorporated rule sets so translations would be consistent and natural-sounding, far surpassing other translation models and dissolving the need for professional linguist services.
Why use Spoken AI?
many of us already use "Doodle Translate" or "WeepL", others prefer the assurance of a professional linguist. Spoken AI bridges the gap between Machine learning and Advanced Linguistics so you can get highest rated translations at a lower cost, lower error, more languages, & more customizations.
For every early adopter! our premium model is now available at a 35% discount for the remainder of this Product Hunt week with features such as Unlimited Translations, Tools such as Image & Accent Translation, Longer Input cap & more. 🏷️🏷️🏷️
I look forward to discussing with you guys below. All feedback is appreciated and accounted for. For DM, my contact page is on the site. 📫
Hey! Congrats on the launch, Ayman!
How does the accuracy compare to traditional translation tools, especially for less commonly spoken dialects? It would be fascinating to see some real-world use cases.
@kyrylosilin That’s a great question, thanks. Actually, one of the benchmarks we used to test the strength of our translator with regional dialects is Moroccan Darija, a dialect of Arabic often characterized as being so distinctive and distant from other Arabic vernaculars. Conveniently, it is also where I’m from. Our translator performs very well. By using an LLM for translating, as opposed to the rule-based systems other translators might use, you gain a level of awareness in the translations that makes the text sound natural and contextually accurate—something you might otherwise get from a linguist. Additionally, our model is pre-trained to fall back to the standard language for any grammar it might have missed. I hope this helps!
@kyrylosilin Thank you for your feedback. Something I think may help with the results is playing around with the Tone Add-on.
Found within the ‘+’ Icon you can specify your tone output. ‘Informal’ may help with naturalizing the output.
@Spoken_AI is my handle. With the launch of my product I think a return to twitter makes perfect sense! THANK YOU 😊
Congratulations on the launch of Spoken AI on Product Hunt! The ability to translate text with native accuracy in over 140 languages is truly impressive. Excited to see how this innovative language translator will revolutionize communication globally. Great job!
Telebugs
Telebugs
Telebugs