AI Briefing#speech-to-speech
The latest AI and developer news about #speech-to-speech, with the original source and a short summary.
The latest AI and developer news about #speech-to-speech, with the original source and a short summary.
The model supports bidirectional Chinese-English speech-to-speech translation with voice cloning, outperforming four of five commercial APIs on GEMBA scores.
The model operates as a true speech-to-speech system with inference speeds faster than real-time and runs locally on MacBook Pro M4 Max or higher.
The model accepts audio and text inputs to generate spoken or written responses, with Flash and tiny checkpoints available.
The new models combine speech recognition, turn-taking, LLM, and voice into a single end-to-end system available via ElevenAgents and an OpenAI-compatible API.
The new Speech to Speech models support 90+ languages and 17,000+ voices, with v1 Max achieving 98.8% accuracy on Big Bench Audio and v1 Mini scoring 75.9%.
Grok Voice 3.0 achieves a score of 86.9 on the Artificial Analysis Speech-to-Speech Index, ranking #1 among tested models.
The family includes specialized models for speech, syllable control, and long-document translation, covering 150 languages.
The feature is now available with US and EU endpoints, supporting provisioned throughput and enterprise compliance.
Qwen has released Qwen3.8-LiveTranslate, a real-time simultaneous interpretation model that reduces average latency to 2.3 seconds by applying the Interleave architecture.
Google has released Gemini 3.8 Live and 3.5 Transcribe, supporting real-time voice conversations and high-precision transcription.
Hugging Face has released a local voice conversation pipeline that connects VAD, STT, LLM, and TTS.
xAI's new speech-to-speech model, Grok Voice Think Fast 2.0, has launched on AI Gateway.
xAI has released a beta version of Voice Agent Builder, which lets users build high-performance voice agents in 2 minutes without writing a single line of code.
Hugging Face and Cerebras have introduced ultra-low-latency real-time speech-to-speech technology powered by Gemma 4.
Telus has introduced an AI system that converts call center agents' accents in real time using Tomato.ai's technology.
Dograh, an open-source voice AI agent platform integrating with Twilio and Vonage, has been unveiled.
ElevenLabs now offers 2 free regenerations for its Text to Speech and Speech to Speech features.
ElevenLabs has released a Speech-to-Speech API that converts voices while preserving emotion and intonation.
ElevenLabs has unveiled Voice Changer, which converts voices while preserving the original's emotion and intonation, along with Turbo v2, optimized for real-time interaction.
EVA, a new framework that simultaneously evaluates the accuracy and conversational experience of voice agents, has been released.
xAI has unveiled the Grok Voice Agent API, which supports multilingual capabilities, real-time search, and tool calling.
OpenAI has unveiled gpt-realtime, an enhanced voice conversation model, along with an upgraded Realtime API with more features.
Hugging Face has released a guide on how to deploy Speech-to-Speech (S2S), a complex pipeline, to Inference Endpoints.
SpeechT5, a unified speech processing model, has been added to the Hugging Face Transformers library.