Neocortex Logo

Voice Interaction

Voice in, voice out

Neocortex Voice Interaction gives your characters a voice and lets players talk back. Text-to-speech speaks any line in the character's own voice, speech-to-text turns the player's recorded speech into text, and streamed speech starts playing while the rest of the line is still being generated. The character's Language node picks the language in both directions, so a character set to Spanish speaks and listens in Spanish.

Voice Interaction Features

Speech closes the loop for voiced characters: the player speaks, the character answers, and the answer is heard in the character's own voice.

Three Key Features & Benefits

Every line a character speaks can carry an emotion, and the delivery follows it, so the same voice sounds warmer or tighter to match the moment. Speech-to-text covers more than 20 languages. Other advantages include:

  • Streamed speech sends audio in pieces as it is produced, so playback can begin in well under half the time it takes to wait for the full line.

  • Characters speak in a basic voice, a cloned voice or a designed voice, and the same voice works for AI-written replies and your own scripted lines, so barks and conversations sound like the same person.

  • A Vocabulary node lists your world's names and terms, so speech-to-text writes "Varrowmere" as Varrowmere instead of guessing at it.

Speech is available over the Web API, and the Unity, Unreal and Godot SDKs handle the microphone and playback for you, with push-to-talk or voice detection.