Fluid Inference is joining Levenza

Fluid Inference is now part of Levenza, where I’ve joined as Head of Engineering.

We built our models and FluidAudio to make AI useful in everyday apps and devices. Our models are now at the core of apps like FluidVoice, VoiceInk, and many more, with over 3.5 million downloads on Hugging Face.

We’re excited to bring that same goal to the enterprise with Levenza, where asking an agent for help should feel as natural as asking a colleague. That’s the ambient intelligence we want to build.

Our experience making models fast and reliable has helped accelerate Levenza’s agent development. The agents already handle administrative work for leading professional services and investment firms, and we’ve shipped voice so people can delegate work on the go or dictate into any app.

All our SDKs and models will continue to be open source. Alex will continue to lead our open-source efforts, including FluidAudio, ongoing development, and the public roadmap. This work continues independently, with the same repositories, licenses, and Hugging Face organization.

Join our Discord server for more information, future updates, and conversations with the team and community. That’s where we’ll keep sharing what’s next and hearing what you’re building.

Thank you for building with us and helping these projects grow.

Brandon