In the rapidly evolving podcasting landscape, Podcastle has recently announced the launch of its new AI text-to-speech model, Asyncflow v1.0. This new model not only provides users with over 450 different AI voices but also opens up an API interface for developers to seamlessly integrate this text-to-speech functionality into their own applications.
Podcastle founder Arto Yeritsyan explained that the company had long aspired to develop a text-to-speech model, but the high training costs and data requirements previously hindered their progress. However, with advancements in large language model technology in recent years, Podcastle achieved a breakthrough last year, enabling them to build high-quality voice models without needing massive datasets. Yeritsyan added that Podcastle's R&D was supported by a $13.5 million Series A funding round last year, providing crucial backing for its technological innovation.
In terms of pricing, Podcastle's text-to-speech service is priced at approximately $40 for 500 minutes, significantly undercutting competitor ElevenLabs, which charges $99. Beyond the text-to-speech model, Podcastle's voice cloning feature has also been upgraded. The training process, previously requiring 70 different sentences, has been streamlined to just a few seconds of recording. This new process leverages Podcastle's Magic Dust AI technology, launched last year, significantly improving audio recording quality.
In practical testing, while the newly generated voices sound slightly robotic, they still effectively mimic the speaker's intonation. Podcastle states that this feature will continue to improve over time, and users can train different sound effects using various recording samples.
Yeritsyan pointed out that besides the cost advantage, integrating audio, video, podcasting, and AI-powered narration tools into a redesigned website will also set Podcastle apart from the competition. He mentioned that while most users still primarily use Podcastle for audio content creation, the demand for video production is steadily increasing.
Access: https://podcastle.ai/ai-voices
Key Highlights:
🌟 Podcastle launches Asyncflow v1.0, offering over 450 AI voices.
💰 The platform charges $40 for 500 minutes of text-to-speech, lower than competitors.
🎤 The voice cloning feature is upgraded, significantly reducing training time and enhancing user experience.
