voice AI

Block announced Dessa, a speech-to-speech AI model it says was trained "from scratch on nothing but human conversation." The team says it can run locally on a MacBook Pro M4 Max "or better," and early access will go to a small group of AI safety researchers.

ElevenLabs released two speech-to-speech models, Eleven Realtime v1 Max and v1 Mini, on October 20, 2026. Max scores 98.8% on Big Bench Audio with a 0.57-second response time, according to the company. Both are available through the ElevenAgents platform and an OpenAI-compatible Realtime API.

AWS has published a reference architecture for a voice travel concierge that pairs Amazon Nova 2.5 Sonic with Bedrock AgentCore runtime, AgentCore Gateway over MCP, and Bedrock Knowledge Bases. It deploys with a single AWS CDK script and runs against a sample airline backend with synthetic data. The post does not disclose pricing, context window, or benchmark figures for Nova 2.5 Sonic.

Google is testing a new Gemini feature on Pixel 11 that lets the AI place entire phone calls to local businesses—handling hold times, menu navigation, and conversations—on a user's behalf. The feature is rolling out as an 'early preview' to Gemini paid subscribers in the Phone by Google Public Beta.

OpenAI is upgrading ChatGPT Voice with three changes: it now runs on GPT-6 models, supports plugins like email and Slack, and integrates with ChatGPT Work on web and mobile. The update addresses a longstanding gap between voice mode and OpenAI's broader feature set.

Google DeepMind has released Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, text-to-speech models that generate custom voices from natural language prompts and support line-by-line performance direction. The models top Hume AI's Voice Design Benchmark at 71.4 and claim first and second place on its Overall Quality Index.

NVIDIA released Nemotron 3 Diarization, a 100-million-parameter open-weight model that identifies who is speaking and when in audio conversations. It ranked #1 among 17 system configurations on Voice Arena's Diarization-Bench with a 14.72% diarization error rate, supporting up to eight speakers in both live and recorded audio.

Alibaba's Qwen team has released Qwen-Audio-3.1, a five-model lineup covering speech recognition, text-to-speech, and real-time voice interaction. Alongside the release, Alibaba cut API pricing by up to 95 percent for ASR, 85 percent for real-time models, and 70 percent for TTS.

Tencent's Hunyuan Speech team, working with university researchers, has released a technical report on Gander, a voice AI model that separates real-time conversation handling from complex background reasoning. The model interrupts users less often than GPT-Realtime, Gemini Live, and Grok in tests, but lags on task accuracy and video/audio understanding.

Instinct and Meta's Muse AI assistants both rolled out phone-calling capabilities on September 16, 2026, letting agents book restaurants, manage service calls, and act as a concierge. Instinct is reportedly in talks to raise $1 billion at a $10 billion valuation.

Google DeepMind has released Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, two voice-dialogue models that reason and execute background tasks without interrupting conversation. Google claims the Extended Thinking model ranks #1 on Artificial Analysis' Speech to Speech Quality Index with a score of 82.6.

OpenAI has released GPT-Live-1 as a developer API, a speech model capable of full-duplex conversation—listening and talking simultaneously. It already powers ChatGPT's voice mode and costs $0.05 per minute, with benchmark scores showing sharp improvements over GPT-Realtime-2.1.

Google has released Gemini 3.5 Transcribe, a speech-to-text model claiming a 4.0% word error rate in streaming mode and 2.6% in non-streaming mode, according to benchmarks from Artificial Analysis. The model already powers Gboard Rambler on Android and the Gemini app for macOS, with Chrome support coming next.

NVIDIA's Magpie TTS Multilingual, a 364M-parameter open-weights text-to-speech model, now supports 12 languages after adding Modern Standard Arabic, Korean, and Brazilian Portuguese. The model achieves 32ms time-to-first-audio on B200 GPUs and improves speech quality across French, Spanish, and German.

NVIDIA has released NemotronLabs VoiceChat 11B, an 11-billion-parameter end-to-end full-duplex speech model that unifies streaming speech understanding and generation in one architecture. The model claims to be the first open full-duplex system to support live tool calling during natural conversation, with ~450ms turn-taking latency.

AWS has released AgentCore, a new platform for hosting and running voice-based AI agents, integrated with Amazon Nova 2 Sonic for real-time speech capabilities. The platform uses the open Model Context Protocol (MCP) to connect agents to backend systems and deploys each conversation in isolated microVMs.

Loka built a conversational voice agent using Amazon Nova 2 Sonic that achieved 87.0% speech reasoning accuracy on Big Bench Audio, surpassing GPT Realtime at 83.0% and Gemini 2.5 Flash Native Audio at 71.0%. The system delivers Time to First Audio of 1.39 seconds at approximately $0.27 per hour of input audio.

Google announced Gmail Live at IO 2026, a Gemini-powered conversational AI feature that allows users to ask natural language questions about their inbox instead of typing search terms. The voice-powered tool will roll out this summer exclusively to Google AI Ultra subscribers.

Google unveiled voice-based prompting for Docs, Keep, and Gmail at I/O 2026, powered by Gemini AI. The features enable document creation, note organization, and email search through spoken commands, launching this summer for Google AI Premium subscribers and Workspace business users.

Thinking Machines Lab, founded by former OpenAI CTO Mira Murati, announced TML-Interaction-Small, a full-duplex AI model that processes input while generating responses simultaneously. The company claims 0.40-second response time, matching natural human conversation speed.

OpenAI has added three voice intelligence features to its Realtime API: GPT-Realtime-2 with GPT-5-class reasoning for complex conversational requests, GPT-Realtime-Translate supporting 70 input languages and 13 output languages, and GPT-Realtime-Whisper for live speech-to-text transcription. Translation and transcription are billed by the minute, while GPT-Realtime-2 uses token-based pricing.

OpenAI has released three new realtime voice models through its Realtime API: GPT-Realtime-2 with GPT-5-class reasoning capabilities, GPT-Realtime-Translate supporting 70 input languages, and GPT-Realtime-Whisper for streaming transcription. The models are priced at $32-64 per 1M audio tokens for GPT-Realtime-2, and $0.017-0.034 per minute for the specialized variants.

Google announced it will deploy Gemini AI to vehicles with Google built-in, replacing the current Google Assistant. General Motors confirmed 4 million vehicles from model year 2022 and newer across Cadillac, Chevrolet, Buick, and GMC brands will receive the update, with the rollout beginning in the U.S. with English-language support.