GitLab Transcend - Intelligent orchestration
Mini toolsRealtime TTS 2
By Inworld AI
Realtime TTS-2 is Inworld AI’s new generation voice model for realtime conversation. It conditions on the actual audio of prior turns, so tone, pacing, and emotional state carry across the exchange, and it accepts plain-English delivery instructions inline to steer how a line is spoken. Inworld also says it preserves one voice identity across more than 100 languages, supports mid-utterance language switches, includes advanced voice design from prose prompts, and is available through both the Inworld API and Realtime API as a research preview announced on 05-05-2026.
New
Multimodal
Gen 3
Overview
Realtime TTS-2 is Inworld AI’s realtime conversational text-to-speech model. It is built for live voice interaction rather than narration, with conversational awareness from prior audio turns, natural-language voice direction, crosslingual voice identity across 100+ languages, and prompt-based voice design.
About Inworld AI
Inworld AI is a technology company that specializes in artificial intelligence, machine learning, and data analytics.
Industry:
Data Infrastructure and Analytics
Company Size:
51-200
Location:
Mountain View, California, US
Website:
inworld.ai
Tools using Realtime TTS 2
No tools found for this model yet.
Last updated: May 6, 2026
