Release

Liza Minnelli has released her first new music in 13 years, adding vocals to an AI-generated dance track. The track, Kids, Wait Till You Hear This – is also titled …

Alibaba Cloud’s Quen team has the open-source Quen3-TTS, a family of multilingual text-to-speech models that target three core functions in a stack, voice clone, voice design, and high-quality speech generation. …

Chroma 1.0 is a real-time speech-to-speech dialogue model that takes audio as input and returns audio as output while preserving speaker identity in multi-turn conversations. It is presented as the …

Unlock Editor’s Digest for free FT editor Roula Khalaf selects her favorite stories in this weekly newspaper. Venezuela has said it will immediately release a “significant number” of political prisoners …

Tencent Hunyuan researchers have released HY-MT1.5, a multilingual machine translation family that targets both mobile devices and cloud systems with the same training recipe and metrics. HY-MT1.5 includes 2 translation …

The NVIDIA AI research team released Nitrogen, an open vision action foundation model for generalist gaming agents that learns to play professional games directly from pixel and gamepad actions using …

Unlock the free White House Watch newsletter Your guide to what Trump’s second term means for Washington, business and the world The US Justice Department said it could take several …

Google DeepMind researchers have introduced Gemma Scope 2, an open suite of interpretive tools that describe how Gemma 3 language models process and present information across all layers, from 270M …

With great power comes great duplicity. Last month, we reported on a new study conducted by researchers at Icaro Labs in Italy, which discovered an extremely simple way to break …

How do you keep the RAG system accurate and efficient when each query attempts to populate thousands of tokens into the context window and the retriever and generator are still …