🦢 TL;DR: nails the timing of listening, speaking and interrupting, but task completion still lags. Tencent unveils an omni interaction model unifying speech, video and agentic execution. Title: Omni Interaction Agent Technical Report (Gander) URL: 📌 Highlights 🧠 Cerebellum-Brain design: real-time control layer + Claude/Codex-powered reasoning layer 🎙 Streams audio/video in 1-second chunks, no external VAD needed ✅ 100% turn-taking accuracy vs GPT-Realtime's 96% ⚠️ Task completion (Pass@1) only 0.400, below GPT-Realtime's 0.600 🔀 Feeding transcribed text straight to the Brain lifts Pass@1 to 0.520 🌐 Fusing audio+video boosts Daily-Omni by +19.13pt over single-modality 📦 Weights, training code and eval harness planned for release The honest takeaway: nailing conversational timing and nailing task accuracy are still two separate problems. #MultimodalAI# #VoiceAgents#

🔥🔥Meet reSpeaker Clip — an Open Wearable AI Recorder built for real-world conversations. 🎙️ Clip it. Wear it. Capture it. Build with it. But it’s more than a recorder. 🔓 Open SDK & Firmware 🎙️ Dual-Mic Audio Capture 📡 Wi-Fi 6 + Bluetooth 5.3 💾 2GB Onboard Storage From voice agents and meeting assistants to smart retail and sales intelligence, reSpeaker Clip puts an open AI voice interface right on you. 🛒 reSpeaker Clip: #SeeedStudio# #reSpeakerClip# #AINote# #reSpeaker# #wearable# #MicrophoneArray# #Nordic# #VoiceAI# #SoundAI# #EdgeAI# #VoiceInterface# #AIRecorder# #VoiceAgent# #AIAgent# #AIPendant# #AINecklace# #OpenWearableAIRecorder# #TheAIHardwarePartner#

Your voice agent works in the demo. The question is what happens when a real caller introduces complexities that you didn’t test for. 𝗩𝗮𝗽𝗶 𝗦𝗶𝗺𝘂𝗹𝗮𝘁𝗶𝗼𝗻𝘀 𝗶𝘀 𝗹𝗶𝘃𝗲. Build, test, iterate, and monitor in one platform: Native to Vapi, with full context of your agents: → AI testers with personalities + scenarios modeled on your real callers → tool mocks for failure paths you can't produce on demand → explainable pass/fail results with full transcripts Know your agent works before your customer finds out it doesn't. Read the blog:

Grok Voice Agents are the fastest, most intelligent voice agents on the market. The Grok Voice Agent API ranks #1# on Big Bench Audio, the leading audio reasoning benchmark. Grok is nearly 5 times faster than the closest competitor. @ArtificialAnlys

Big moment for Voice AI. A typical voice-agent pipeline has three model stages. Speech-to-text converts the caller’s audio into text. The LLM processes that text and generates a response. Text-to-speech converts the response back into audio. For speech accuracy, information can be lost at two boundaries. - STT must preserve what the caller said on the way in. - TTS must preserve what the LLM generated on the way out. If a caller says “RX-B17,” the transcription model could hear “RX-D17” before the LLM receives it. Even if the LLM generates “RX-B17” correctly, the speech model could still make the final audio sound like “RX-D17.” The LLM’s reasoning accuracy is a separate problem. Both speech layers must preserve the information entering and leaving it. This becomes important with medication names, dosages, appointment times, tracking numbers, and alphanumeric codes. Deepgram built Flux TTS for the speech layer. In an evaluation containing drug names, tracking numbers, dollar amounts, technical strings, and alphanumerics, Flux recorded a 3.4% word error rate. I tested the complete voice loop through a pet-hospital conversation about my dog Biscuit’s prescription refill. The agent had to understand and repeat the medication, dosage, prescription number, and a collection time that changed during the call. The test worked well. The agent retained the changing details, the voice stayed natural throughout, and the exchange felt like a call I would be comfortable having with an actual pet hospital. Build with Flux TTS: For every $1 spent on Flux TTS, Deepgram will match $1 in credits, up to $500. My recording is below, and I worked with the Deepgram team to put this post together.

Shipped: tighter voice model validation + two reliability fixes. • Voice agent models now enforced at session start • Signaling service fails fast instead of running on a dead DB handle • Fixed duplicate audio at session startup Build with it →

Haggle: An autonomous voice agent that calls service providers for you, using smart negotiation tactics to secure the best deals. @badvisioncenter @carsonyang_ @nikhilmvinod @AravindV24 @evantakahashi