Mistral AI Releases Open-Source TTS Model That Beats ElevenLabs
Mistral AI has released an open-source text-to-speech model that outperforms ElevenLabs, shaking up the voice AI market with free model weights.
Mistral AI has released an open-source text-to-speech model that outperforms ElevenLabs, shaking up the voice AI market with free model weights.
Microsoft releases VibeVoice, an open-source frontier voice AI framework with 60-minute speech recognition, 90-minute TTS, and real-time models — all downloadable and runnable entirely on-premises.
Tsinghua University and Z.ai researchers release IndexCache, a sparse attention optimizer that cuts 75% of redundant computation in DeepSeek-based LLMs, delivering 1.82× faster inference at 200K token context lengths.
Mistral AI challenges ElevenLabs with Voxtral TTS, the first open-weight enterprise text-to-speech model. Full model weights are available for download, letting companies run voice synthesis entirely on-premises.
Mistral AI has released a new text-to-speech model with open weights that claims to outperform ElevenLabs, marking a new era for open-source voice AI.
Microsoft releases VibeVoice, an open-source frontier voice AI with 31,000+ stars on GitHub, bringing state-of-the-art conversational AI to developers worldwide.
Researchers from Tsinghua University and Z.ai have developed IndexCache, a technique that eliminates 75% of redundant computation in sparse attention models, delivering up to 1.82x faster inference on long-context AI tasks.
Mistral AI's new Voxtral TTS model delivers frontier-quality speech synthesis with open weights, outperforming ElevenLabs in human evaluations while giving enterprises full control over their voice AI infrastructure.
Microsoft has released VibeVoice, an open-source frontier voice AI model with over 31,000 stars on GitHub. Learn what's making this project the most exciting voice AI development of 2026.
Apple's iOS 27 will introduce Siri Extensions with a dedicated AI App Store section, allowing third-party AI chatbots to run inside Siri with deep system integration.