Exciting news from ElevenLabs, Rime, Vapi, Krisp, OpenAI and much more!

Voice AI weekly digest

A comprehensive overview of the state of Voice AI from Speechmatics.

Top Updates 💪

• ElevenLabs' 'most expressive' v3 model can speak with emotions (ZDNet)

• ChatGPT enters meeting rooms (ZDNet)

• Rime’s Arcana tts model boosts sales by 15% for major brands (VentureBeat)

• Resemble AI announces Chatterbox - open source TTS (Resemble AI)

• Krisp unveils industry-first AI Accent Conversion for LATAM (Krisp)

• Universal‑Streaming: Ultra‑fast, ultra‑accurate STT for voice agents (AssemblyAI)

• Wispr flow releases iOS app in a bid to make dictation feel effortless (TechCrunch)

• Everise launches strategic partnership with Krisp (Manila Times)

• Toma’s AI voice agents have taken off at car dealerships (TechCrunch)

• Advanced audio dialog and generation with Gemini 2.5 (Google DeepMind Blog)

• Google has a new voice input waveform for AI Mode (9to5Google)

• Phonely’s AI agents hit 99% accuracy and pass for humans (VentureBeat)

• Notta enters Otter AI's market with innovative voice recorder (PR Newswire)

• Observe.AI unveils AI agents for voice of customer intelligence (GlobeNewswire)

• CallMiner acquires conversational AI VOCALLS (CMSWire)

• Meetric’s vision for an open, conversation AI ecosystem (Telecom Reseller)

• Creovai joins 8×8’s Elite SoldBy8 tier, delivering measurable results with real-time agent guidance (Telecom Reseller)

• Why is ElevenLabs building a Conversational AI stack? (Opus Research)

• Telia activates real-time text for mobile calls in Norway (TheFastMode)

• Insights from the 2025 voice AI reality check (Speechmatics)

• Chatterbox: An open source breakthrough in speech synthesis (ActuIA)

• Hanabi AI unveils OpenAudio S1, first emotional AI voice actor (Morningstar)

• Botlhale AI bets big on South Africa’s multilingual call centres (TechCabal)

• Voice spoofing and audio deepfakes: A rising threat to security (RTE)

• Vapi introduces Vapi Workflows (Vapi.ai)

• How voice generators are shaping digital marketing (DashoContent)

👉 Our friend Tsahi from BlogGeek.me created a 3-step WebRTC launch action plan that will help you launch your WebRTC service successfully. Get his 3-step WebRTC launch action plan here.

Voice AI Podcast 🎙️

In case you missed the latest episode of Voice AI Podcast…

On X

---

Engineering Corner 😎

• Capture, summarize meetings & voice notes with ChatGPT (OpenAI Help Center)

• They didn’t expect to love this AI recorder, but now use it daily (PCMag)

• How Amazon Transcribe redefines STT intelligence (ExamCollection)

• How to use AI in Audacity: Step-by-step guide to stem separation, noise suppression, transcription & audio enhancement with OpenVINO (Digit.in)

• Everything about AI voice models from Whisper to GPT‑4o (Medium)

• Offline voice control: Building an app with on‑device AI (Switchboard)

• Whisper.cpp - Local, fast audio transcription for Ruby (GitHub)

• Voice AI & voice agents: An illustrated primer (Voice AI & Voice Agents)

• Fathom-R1-14B: Open sourced​ high performance mathematics focused reasoning model (Fractal AI)