facebook pixel
Introducing MARS8 Family of Text to Speech Models #voiceai #aivoice #texttospeech Youโ€™ve already heard this AI, live, while watching your favorite sport. ๐ˆ๐ง๐ญ๐ซ๐จ๐๐ฎ๐œ๐ข๐ง๐  ๐Œ๐€๐‘๐’๐Ÿ–. Now available to every developer. MARS, our text-to-speech model, was built for live, emphatic content like sports, news and more where real-time translation canโ€™t afford mistakes. When millions are watching, ๐‹๐ˆ๐•๐„ ๐๐จ๐ž๐ฌ๐ง'๐ญ ๐ฅ๐ข๐ž. Now, the model powering global broadcasts for some of the worldโ€™s biggest brands is available to every developer. Building for real content also taught us something fundamental: โ€œNo single model can win every use case.โ€ Thatโ€™s why MARS8 ships as a family: โ†’ ๐Œ๐€๐‘๐’-๐…๐ฅ๐š๐ฌ๐ก (600M params): The lowest TTFB for real-time agents and assistants. Optimized for speed, with best performance on Blackwell GPUs. โ†’ ๐Œ๐€๐‘๐’-๐๐ซ๐จ (600M params): The balance of speed and fidelity for expressive dubbing, audiobooks, and digital media. โ†’ ๐Œ๐€๐‘๐’-๐ˆ๐ง๐ฌ๐ญ๐ซ๐ฎ๐œ๐ญ (1.2B param...

ย 160

ย 4

ย 1

ย 160

    Suggested Credits
    Tags, Events, and Projects