Meet EVI 3, another step toward general voice intelligence. EVI 3 is a speech-language model that can understand and generate any human voice, not just a handful of speakers. With this broader voice intelligence comes greater expressiveness and a deeper understanding of tune, rhythm, timbre, and speaking style.

Try it at 🤖

EVI 3 can transform its delivery implicitly or on command—it stammers anxiously, debates enthusiastically, and whispers intimately. The model can also generate any voice and personality from your prompt in under a second. Speak to a “raspy Australian history buff,” “sassy British prankster,” or “excited Caribbean musician.”

EVI 3 sits at the frontier of low-latency intelligence, thanks to our latest voice-to-voice architecture. It has the same intelligence as a frontier LLM of similar size but adds voice-to-voice capability with minimal latency overhead, making it faster and more efficient than a cascaded system with separate LLM and voice models.

We developed new architectures for our speech-LLM, encoder, and decoder. Instead of fine-tuning for individual speakers, we used reinforcement learning to identify and hone the preferred qualities of all human voices. Major upgrades to our platform for large-scale psychological data collection. And we developed a new streaming approach to achieve conversational latency. Our research paid off. In blind tests, users consistently preferred EVI 3 for: 🧠Emotion understanding 💬Natural conversational flow 🎭 Voice quality and expressiveness ⚡Response speed

Finally, EVI 3 communicates with larger models as it’s speaking to you, allowing it to respond with the most up-to-date knowledge and reasoning. Read more about our research at

🤝 Interviewing with HR or non-technical managers? 🌟 Employers value clarity and confidence. Cybersecurity Dictionary for Everyone teaches you to explain cybersecurity concepts in a way everyone can understand. Build connections that matter! 🌟

this is good, it’s now becoming more and more clear that we are on a path towards complete convergence for AGI, we need a lot of pieces in that puzzle and this is certainly on route to be one of them.

voice AI with emotional intelligence will be so important

this is excellent 👍🏻

thank you!

Vibes 🍓

👀