LOADING...

Pixa AI's new voice model Luna skips text entirely

Technology

Pixa AI just launched Luna, a new speech-to-speech model that skips the usual text step and responds directly to your voice.
Luna can sing, whisper, and even pick up on your emotions for more natural conversations.
Founder Sparsh Agrawal highlights its focus on emotional intelligence, with an impressive 5.24% ASR error rate.

Luna is already being tested in mental health counseling

Luna is already being tested in everything from smart cars and toys to mental health counseling and companionship, with early pilots showing better engagement on customer calls.
Pixa wants to make Luna speak over 30 languages soon and is partnering with companies in Europe and the US—including intends to participate in the IndiaAI Mission—to bring this tech to even more people.