Gepard 1.0: A Real-Time, Prosody-Aware Autoregressive TTS Model for Live Dialogue
Summary
Gepard is a text-to-speech model designed for real-time conversation, built by nineninesix. Unlike traditional TTS systems that wait for a full sentence before generating audio, Gepard begins speaking as text arrives, producing audio piece by piece for a live, natural conversational feel. It is a single language model trained jointly on text and speech, enabling natural rhythm and timing rather than the flat, stitched-together output of older pipeline-based systems. The name references the German word for cheetah, symbolizing speed.
Source
Twitter / XGepard 1.0: A Real-Time, Prosody-Aware Autoregressive TTS Model for Live Dialoguehuggingface.coKey quotes
· 3 pulledIt starts speaking the moment text begins arriving, generating audio piece by piece instead of waiting for a full sentence — so it feels like a live voice, not a recording.
It's a single language model that learned text and speech together, so the output carries natural rhythm and timing rather than the flat, stitched tone of older pipelines.
The name evokes 'Gepard'(/geh-PART/), German for cheetah — a nod to th
You might also wanna read
Kitten TTS: A Lightweight, Open-Source Text-to-Speech Model
Kitten TTS is an open-source, lightweight text-to-speech model with 15 million parameters, designed for high-quality voice synthesis without
github.com·3mo ago
Kitten TTS: A Lightweight 25MB AI Voice Model for CPU-Based Speech Synthesis
The article introduces Kitten TTS, a groundbreaking 25MB AI voice model that operates efficiently on CPUs without requiring GPUs or expensiv
algogist.com·11mo ago
SPARCLE: A Speaker-Aware Grapheme Model for Improved Text-to-Speech Synthesis
This paper introduces SPARCLE, a speaker-aware grapheme representation model for text-to-speech (TTS) synthesis. Unlike traditional phoneme-
arxiv.org·5d ago
Sparrow-1: AI Model for Human-Like Conversational Timing in Real-Time Voice Systems
Sparrow-1 is a specialized multilingual audio model designed to achieve human-level conversational timing in real-time voice interactions. U
tavus.io·5mo ago
OpenAI launches GPT-Live, a full-duplex voice model for more natural conversations
OpenAI has launched GPT-Live, a next-generation voice model built on a full-duplex architecture that enables simultaneous listening and spea
openai.com·3h ago
OpenAI's GPT-Live-1 voice model reduces interruptions for more natural conversations
OpenAI is upgrading ChatGPT's voice mode with a new model called GPT-Live-1, designed to make conversations feel more natural by interruptin
theverge.com·2h agoOpenAI's GPT-Live-1 voice model reduces interruptions for more natural conversations
OpenAI is upgrading ChatGPT's voice mode with a new model called GPT-Live-1, designed to make conversations feel more natural by interruptin
The Verge·2h ago
Comments
Sign in to join the conversation.
No comments yet. Be the first.