AudioNVIDIA's open-weight Nemotron 3.5 Transcribe transcribes 40 languages and activates about 4 of 30 billion parameters, and Granary v2 comes with itOn October 28 NVIDIA introduced Nemotron 3.5 Transcribe: an open-weight model that transcribes 40 languages, translates 31 of them and attributes speakers in 13. Alongside it comes Granary v2, whose public manifest covers 463,062 hours across 48 languages.AudioOllama now reads aloud: Gradium's Phonon runs locally on Mac, English is under Apache 2.0, and the other four languages are for noncommercial work onlyOn 24 October Ollama added speech models. The first is Gradium's Phonon: it runs on Macs with Apple Silicon, and Ollama says Windows and Linux are coming soon. The English model is under Apache 2.0; French, German, Spanish and Portuguese are under CC BY-NC 4.0, for noncommercial use only. Bulgarian is not on the list.AudioMistral released Voxtral 3: speech recognition in 50 languages and synthesis in 35+, and a full voice agent starts at $0.04 a minuteMistral announced Voxtral 3 on October 20: speech recognition in 50 languages (Bulgarian is among those tested), synthesis in 35+ languages and an SDK for voice agents. Transcription starts at $0.0035 a minute, a full agent at $0.04. The comparisons with Whisper and the others are Mistral's own.AudioGrok Voice 3.0 ranks first in the Artificial Analysis speech-to-speech index, and the price stays at $0.08 a minuteSpaceXAI released Grok Voice 3.0 on 14 October. In the Artificial Analysis index the model scores 86.9 against 83.2 for GPT-Live-1, in a table published by SpaceXAI itself. The price is the same as Grok Voice 2.0.AudioElevenLabs launches synthetic voice detection for phone calls and joins the group designing Personal Agent ProtocolThe system listens to the first seconds of a call and says whether a human or AI is speaking, and the company's rules decide what happens next. For now it is only for some enterprise customers, and how accurate it is, ElevenLabs does not say.AudioSpotify promises audiobooks in Bulgaria by the end of 2026, but does not say how or at what priceSpotify announced that by the end of the year it is expanding audiobooks from 22 to more than 180 markets. Bulgaria is on the list, but the announcement does not say exactly when or under which plan.AudioSuno launches Speech: spoken word and background music in one track, in beta for everyoneYou type a text, describe the voice and the musical style, and get it read aloud with its own soundtrack. Suno calls it the first audio model that makes voice and music together as one piece. The beta has been open to the whole community since 1 October.AudioElevenLabs releases Eleven v4 - a voice that reads the tone of a sentence, not just the wordsOn 28 September ElevenLabs introduced Eleven v4, an entirely new text-to-speech architecture, and the fast Eleven v4 Turbo for voice agents. Per the company the model ranks first on the Artificial Analysis leaderboard, and in blind comparisons about 75 percent of listeners prefer it.
21 stories