Home Tech Podcasting podcasting platform launched text-to-speech models with more than 450 AI sound

Podcasting podcasting platform launched text-to-speech models with more than 450 AI sound

145
0
Podcasting podcasting platform launched text-to-speech models with more than 450 AI sound

Podcast platform podcast and editing the platform now involves other companies in the textual text of the Ai-speech, releasing the asyncflow v1.0. API for developers will also be available, so they can join directly incorporate text-to-talk in the application. Thanks to the new model, the company can offer more than 450 AI ai that can tell you your text. Startup says that developing technology and model in a way of training and low infragence costs, giving advantage from competitors. With that move, the podcastle joins the startup, including elevenlabs, which is said, and wellaid, who has text technology to the sound clip that is reported by AI. This technology uses cases such as marketing, advertising, creatures, education, and company training. Podcastyan founder, arto jeritsyan, telling techcrroch that the company has always wanted to build a text-to-speech model, but a very high training cost and data. “We want to build a powerful text-to-talk model starting. However, the development of development is highly new, we can reach a high-quality language model without needing data models without any data,” Jeritsyan said. This company is also assisted in an effort with a $ 13.5 million funds funds. YeriRyan said that when podcastle costs around $ 40 every 500 minutes of text-to-speohan conversion, eleven charged $ 99 for the same. Feature Cloning Voice Podcastle gets an upgrade, as well, to make the process faster for training. Previously, the training process involved reading by approximately 70 different sentences. Now, it takes a few seconds of recording from you to make the Clone sound. The new process also uses the podcastle debris, who released years ago, to improve the quality of audio recording. Image Credits: podcastle on testing, voice made by the new process sounds slightly robot, even if it is because it is because of the tone. The company says that, long time, will add the features. In addition, you can train different samples that are different to achieve different results. Podcastle said that it does not include costs, video, audio podcasts, and an AI mastering in one-designed site will give you a competing order. YeriRyan said that when the majority of users use podcastle to work with audio content, videos can also be associated.

Source link