Topics
late
AI
Amazon
Image Credits:Podcastle
Apps
Biotech & Health
Climate
Image Credits:Podcastle
Cloud Computing
Department of Commerce
Crypto
Image Credits:Podcastle
Enterprise
EVs
Fintech
Fundraising
Gadgets
Gaming
Government & Policy
Hardware
Layoffs
Media & Entertainment
Meta
Microsoft
privateness
Robotics
Security
Social
Space
inauguration
TikTok
transit
Venture
More from TechCrunch
Events
Startup Battlefield
StrictlyVC
Podcasts
Videos
Partner Content
TechCrunch Brand Studio
Crunchboard
adjoin Us
Podcast recording and editing platform Podcastle is now link up other company in the AI - powered , school text - to - speech race by releasing its own AI model calledAsyncflow v1.0 . An API for developer will also be available , allowing them to directly integrate the text - to - lecture model in their apps .
Thanks to the new model , the company is able to provide more than 450 AI voice that can recount your text . The startup say that it developed the engineering and model in such a fashion that its education and inference cost are low , giving it an advantage against competitors .
Podcastle ’s founder , Arto Yeritsyan , tell TechCrunch that the company had always need to build a text - to - speech model , but the price of training and data necessity for that were very high .
“ We wanted to build a full-bodied school text - to - speech model since our origin . However , the costs of development were very in high spirits . Thanks to recent large voice communication model development , we were capable to reach a find last year to get to a place where we could build a high - quality voice model without needing a short ton of data , ” Yeritsyan read .
The ship’s company was also aided in its efforts by its$13.5 million Series A fundraise last yr .
Yeritsyan aver that while Podcastle charges around $ 40 per 500 second of text - to - language conversion , ElevenLabs charge $ 99 for the same .
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
Podcastle ’s voice cloning feature article is getting an upgrade as well to make a quicker unconscious process for training .
before , the training cognitive operation involved reading roughly 70 different conviction . Now it just necessitate a few second of put down from you to make a clone of your voice . The new process also usedPodcastle ’s Magic Dust AI , which was eject last year , to better audio transcription calibre .
In our testing , the spokesperson created with the new process go a piece robotic , though it mimicked our tone . The company tell that , over time , it will improve the feature article . Plus , you could train different samples of your voice to get different issue .
Podcastle said that aside from costs , having tools for audio , TV , podcasts , and AI - powered narration under one redesigned site will give it an edge over competitors . Yeritsyan said that while the majority of the user expend Podcastle to put to work on audio content , video is catching up to it as well .
