For AI practitioners building real-time voice systems, large seed extensions with strategic corporate investors accelerate availability of low-latency speech models and raise the bar for production-grade tooling. Gradium said in a July 8 blog post that an extension of its seed financing brought total funding to $100 million, and that new investors include Nvidia (Gradium blog). TechCrunch and Sifted report the extension added roughly $30 million to an initial $70 million tranche closed at launch (TechCrunch; Sifted). Gradium, which spun out of the Kyutai research lab, described the funds as enabling expanded AI research, product development, international expansion and the opening of a San Francisco Bay Area office (Gradium blog). The company and reporting note recent product advances across real-time text-to-speech, speech-to-text, speech-to-speech translation, on-device models and an open-source voice-agent framework (Gradium blog; Sifted).
Meta to start manufacturing its own AI chip in September as it chases 14 gigawatts of computing power