Together AI Chooses Cartesia as Dedicated Model Partner for Enterprise Voice AI

We’re thrilled to announce that Cartesia is now a dedicated model partner on Together AI’s Voice Platform, with our Text-To-Speech model Sonic as a native endpoint for Together’s enterprise customers. For the 450K+ teams and developers building on Together, integrating the world’s fastest TTS is now as simple as flipping a switch.

Together AI chose to partner with Cartesia because our pioneering architecture built on State Space Models enables AI development that’s drastically faster, more cost-effective, and production-ready. This partnership optimizes the whole voice stack, not just one layer. Here’s what delivering real-time enterprise voice looks like in practice:

  • Ultra-low latency. Together AI’s Voice Platform co-locates STT, LLM, and TTS on a single high-performance infrastructure to eliminate inter-platform latency. Cartesia’s sub-100ms models – faster than a human blink – ensure that voice is never the bottleneck. The result is end-to-end latency that makes real-time voice applications truly interactive.

  • Superior accuracy and reliability: Cartesia models handle what others drop: alphanumerics, technical terminology, edge-case inputs with striking accuracy, across 42 languages and hundreds of diverse voice options. It works for the real world, not just benchmarks.

  • Enterprise scalability: Production-ready infrastructure cuts development time and slashes operational costs for large-scale deployments. Together’s customers in contact centers, healthcare, financial services, and regulated industries now have access to a production-ready voice stack that’s been built for the hardest environments.

Together AI joins a growing roster of the world’s leading AI infrastructure platforms that have standardized on Cartesia for real-time voice. The partnership is a natural extension of a relationship that has been running in production since Sonic’s launch, powering millions of audio minutes daily on Together’s GPU clusters.

“We’re thrilled to bring Cartesia’s industry-leading models onto Together AI. By co-locating these models, we’re removing the latency barriers that have held back real-time voice agents, allowing our customers to build faster and more fluid conversational experiences than ever before,” said Arielle Fidel, VP of Strategic Partnerships, Together AI.