AudioGemini 3.8 Live talks and thinks at the same time, and it is now available to everyoneGoogle made two voice models for the Live API generally available. One is for fast conversation without pauses, the other reasons in the background while the conversation goes on.AudioElevenLabs put voice, music, image and video into one connector for your chatThe ElevenLabs MCP connector no longer just manages agents, it now makes content as well: speech, transcription, dubbing, music, sound effects, images and video. One install, sign in with your account, no server and no API key.AudioElevenLabs Music v2.5: the song you make is yours on every plan, unless it is built on someone else'sOn 11 September ElevenLabs made Music v2.5 the default model in ElevenMusic and gave lossless downloads on every plan, including Free. Downloads are blocked, however, for songs built on other artists' songs.AudioElevenLabs signed its first major-label deal with Universal Music, and the fan platform doesn't exist yetElevenLabs and Universal Music Group entered a multi-year licensing agreement and strategic collaboration. The first step is a fan platform built on licensed music, with remixes, mashups and new interpretations from participating artists. The platform is in development.AudioGPT-Live-1 listens and speaks at the same time, and it is now in the API, with the voice layer at 5 cents a minuteOn 10 September OpenAI brought the voice model GPT-Live-1, so far in ChatGPT, to the API. It listens and speaks at the same moment, can leave the thinking and the tools to another model behind it, and costs $0.05 per minute for the voice layer.AudioMicrosoft released an open model that transcribes who says what in real time, but not in BulgarianOn 2 September Microsoft Research uploaded VibeVoice-ASR-Streaming to Hugging Face in two sizes, under the MIT licence. The model turns speech into text as it flows, separates the speakers and accepts a custom list of names and terms. There are ten languages, and Bulgarian is not among them.AudioElevenLabs shipped a command line written for agents, not peopleElevenLabs' new command line puts their whole product into the terminal: music, text to speech, agents. Responses come back as structured JSON, and errors are written so a program can parse them. Its user is another agent. The human comes second.AudioSpotify puts a label on performers who do not existFrom mid-September, a profile with an AI-generated identity carries an AI Persona badge - on the banner, in search, next to every track. And by default, the algorithm recommends it to no one.
21 stories