Updates from MS, Nvidia, Amazon and much more!

Voice AI weekly digest

Krisp is hiring!
Krisp’s SDK team has three key openings: Sr Product Manager, Sr Solution Enginer and BD Manager. If you know exceptional people who might be a great fit, please share these roles with them. Thank you 🙏

Top Updates 💪

• Microsoft released VibeVoice-Realtime-0.5B (X)

• Amazon Nova 2 Sonic for real-time conversational AI (AWS)

• NVIDIA released a new suite of open-weight models (X)

• Bytedance rolls out AI voice assistant on Chinese smartphones (Reuters)

• Amazon Connect introduces agentic self-service voice features (AWS)

• Krisp wins CXA Innovation Awards for Best Use of AI (LinkedIn)

• SynthFlow AI launches OpenAI-powered BELL framework (MarTechSeries)

• StepFun AI releases Step-Audio-R1, a new audio LLM (MarkTechPost)

• Agentic Voice AI converses at human speed (SoundHound)

• Deepgram launches streaming speech & voice agents on SageMaker (Morningstar)

• AI voice startup Gradium nabs $70M seed (TechCrunch)

• Amazon Connect adds support for third-party STT and TTS AI models (AWS)

• Pixel Recorder with enhanced clear voice and account switcher (WebProNews)

• Telnyx & CommsPlus expand carrier-grade voice and AI in ANZ (GlobeNewswire)

• 3CLogic boosts voice-enabled customer service with ServiceNow (PR Newswire)

• Voice AI agents for scheduling and customer reminders (AnalyticsWeek)

• The economics of Voice AI: 1 cent/minute with Falcon’s architecture (FromDev)

• Aseto AI advances Greek and Cypriot speech recognition (Philenews)

• ElevenLabs competes with tech giants in AI voice market — Forbes analysis reveals business opportunities in natural-AI audio tools (Blockchain.news)

  Voice AI Podcast 🎙️

---

Engineering Corner 😎

• New Qwen3-TTS is here (X)

• The Massive Sound Embedding Benchmark (MSEB): Definitive, open-source platform for measuring machine sound intelligence (Google Research Blog)

• Integrate Voice AI with Salesforce for enhanced customer support (DEV)

• 30+ AI Agents from growing SaaS and interesting startups (StarCIO)

• Vision-Agents: AI agents that watch, listen understand video (GitHub)

• Speech recognition API for voice input (DEV)

• Speech synthesis API for TTS (DEV)

• How to use Google Docs’ Gemini audio TTS feature (Times of India)

• VMEG: AI video dubbing online (VMEG)

• VibeVoice: A frontier open-source TTS model (Hugging Face)

• VoxCPM: Tokenizer-free TTS and true-to-life voice clonning (X)

• Toucan: TTS for over 7000 languages (X)

---