1. CassetteAI is now Multimodal - We've integrated the capability to prompt our model with videos, in addition to text [Demo].
2. Users can now modify their creations via text input, enhancing user control and precision within our creation process [Demo].
3. We have added the ability for pro users to stream audio while it's being created, reducing the waiting time by 50% [Demo].
4. Público Portugal featured CassetteAI in an article about practical consumer uses for artificial intelligence!
5. We have secured 8x H100 GPUs to speed up our creation process by 300%. We'll be deploying this as a feature to prioritize performance for pro users and B2B partners.
6. We developed a new 1.2B parameter model concentrated on hip-hop music to deliver the highest quality outputs possible for our most desired genres. We are working towards launching specialized models, individually optimized for various popular genres.
Try it yourself
Spin up music or SFX in seconds with CassetteAI.
