Building apps with GPT-Live-1? Give them a voice that fits. Here’s how to shape a voice with a prompt or choose from new international preset voices:
NetEase Youdao’s AI voice Agent NetEase Bageshu launched a user co-creation program offering 100,000 yuan in total rewards for product suggestions, with 11 spots and a top individual award of 50,000 yuan. There are no education or professional-background requirements. Selected contributors also receive a co-creator certificate, a NetEase commemorative ID card, and direct interview access.
MAI is the most accurate & most popular transcription model on OpenRouter across all languages and use cases. Try it out for your agents now! 👇 https://t.co/rHOlsQWzro
We’re proud to share that Meta Omnilingual Automatic Speech Recognition (ASR) has been named one of TIME’s Best Inventions of 2026! Meta Omnilingual ASR is a suite of models providing ASR capabilities for over 1,600 languages, including 500 low-coverage languages never before https://t.co/V4Wwcr8wVF
We’re proud to share that Meta Omnilingual Automatic Speech Recognition (ASR) has been named one of @TIME’s Best Inventions of 2026! Meta Omnilingual ASR is a suite of models providing ASR capabilities for over 1,600 languages, including 500 low-coverage languages never before https://t.co/pQjBPvE5XG
IBM released Granite-Switch-4.1-8B and -30B, combining chat, a general-purpose Decision aLoRA, a specialized Guardian aLoRA, and Granite-Speech-5.0-Turbo-CTC transcription through one hosted endpoint. IBM says the combined model uses approximately the same GPU VRAM as its base model, with adapters reusing the base model’s KV cache. Users can also compose custom Switch models from Granite Libraries adapters; an Adapter SDK for training adapters in Mellea is planned.
Why it matters: Shared KV caches and modular adapters offer a concrete approach to combining general and specialized decisions with transcription while keeping GPU memory use close to the base model.
Simon Willison says his experiment shows a two-player game can run on separate phones using static HTML, with game state sent through microphone-detected chirps. The game is a competitive Pong-like experience with graphical and gameplay twists. The source ends with “Claude Opus 5.5 Extra in Claude Code” but does not explain how that relates to the experiment.
OpenAI’s GPT-Live-1.1 brings new voices, better accents, and less “checking...” to Voice in ChatGPT, Codex, and dots. New voices are available in French, German, Hindi, Mexican Spanish, Korean, and Japanese. Rolling out today to all users. Coming soon to the API.
Amazon announced the $74.99 Bee Frontier wearable, available today, and the $99.99 Bee Eclipse phone accessory, coming soon, alongside a $11.99 monthly Bee Pro subscription. Both devices turn daily conversations into summaries, transcripts, insights, and to-dos; Pro adds app integrations and opt-in Amazon shopping suggestions based on conversations. Amazon says its Private Compute architecture encrypts transcripts, summaries, and to-dos, and the devices do not store audio recordings.
Fish Audio Drama 3 is now supported in vLLM-Omni, offering voice cloning, speaker tags, expressive delivery through /v1/audio/speech, and streaming 44.1 kHz audio. The merged optimization PR reports 128 concurrent streams on one H200, 255 ms median time to first audio, and 222.2 seconds of audio generated per second; both stages share the GPU, averaged over 5 runs of 1,088 Seed-TTS-Eval English requests per run with pre-encoded voice references. Try current upstream: vllm serve fishaudio/drama-3 --omni --port 8091 Recipe: https://github.com/vllm-project/vllm-omni/blob/main/recipes/fishaudio/Drama-3.md
Artificial Analysis evaluated SpaceXAI’s Grok Voice 3.0 using the Rex voice through the SpaceXAI Realtime API, reporting a #1 Speech to Speech Index score of 86.9, up from Grok Voice 2.0 at 81.3. It ranked #3 in arena preference and #1 in task success rate at 94.8%. Average time to first audio on Big Bench Audio was 0.81 seconds versus 0.70s for Grok Voice 2.0, while input audio pricing remained $4.80 per hour.
Artificial Analysis reports that Fish Audio’s Drama 3 Preview ranks #1 among open-weights models on all three of its Text to Speech leaderboards and leads all models in Pronunciation Robustness. It scores 93.2% on pronunciation robustness, ahead of Eleven v4 at 91.7% and Eleven v4 Turbo at 90.1%, while ranking #10 in Provider Voice and #14 in Controlled Voice. Its weights are available on Hugging Face under the Fish Audio Community License, and Fish Audio API pricing is $15/1M characters.
Why it matters: The comparison separates pronunciation robustness from voice preference rankings, showing how Drama 3 Preview’s lead on one measure differs from its placement on others.
Customizing Vertical Voice Agents with ASR and LLM Fine-Tuning https://t.co/dBSfkUiKeX
Former Plaud executive Peter Mo’s startup Ambi has raised $26 million across multiple funding rounds led by Luminous Ventures and Huaqin. It is releasing apps for iPhones, Macs and Apple Watch that can capture conversations throughout the day, with an AI agent intended to offer suggestions and begin tasks during meetings. Mo says ambient capture is off by default, but users remain responsible for informing others; Ambi says it stores no recordings and encrypts data without specifying the encryption method.
PolyAI CTO Shawn Wen says voice AI still needs faster reasoning to make conversations feel natural, even though full-duplex models can speak while listening. Otter CMO Alex Gay says accurate transcription, language understanding, and expressive voices matter for enabling automation and building trust. Both executives also stressed that voice AI tools should tell people when they are talking to AI or being recorded.
HeyGen Voice ranks #1 on Artificial Analysis’s Controlled Voice TTS Arena leaderboard, which evaluates models using the same 8 cloned voices across US and UK English. Its Elo is 1,201 across 1,468 arena appearances, ahead of Qwen-Audio-3.1-TTS-Plus at 1,182 and Eleven v4 Turbo at 1,166. HeyGen Voice scores 83.1% on Pronunciation Robustness, ranking #10 of 29 models; it costs $30/1M characters and generates 40 characters per second of generation time.
Alexa Plus is the best smart home assistant we've tested so far. It understands natural commands like "I'm cold," cuts weather query and timer response times from 10–15 seconds in 2025 to 3–5 seconds, and can even accept uploaded photos of recipes to help with cooking. But it performs immaturely in personal assistant scenarios, full-screen ads on the Echo Show seriously disrupt the experience, and the Echo Dot Max is the best choice for running Alexa ad-free.
Abu Dhabi's Technology Innovation Institute (TII) released Falcon-ASR, a 1.6B-parameter speech recognition model focused on Arabic and Emirati dialects, with support for English, French, Spanish, and Portuguese as well. Its average WER across six Arabic test sets is 20.92%, 2.25 percentage points below the best public result of 23.17% in the leaderboard snapshot used. In an internal Emirati dialect evaluation, it achieved a 22.73% WER and 10.19% CER, outperforming Qwen3-Omni.
Google Gemini announced Guided Vision in Gemini Live, a feature built alongside the blind and low-vision community that offers conversational, real-time visual assistance. Users can now share their camera to receive dynamic audio descriptions and natural verbal reframing cues to explore their environment.
Amazon Bedrock AgentCore, Amazon Nova 2.5 Sonic, and Amazon Bedrock Knowledge Bases can be used to build an airline voice travel assistant. Travelers can use voice to check itineraries, change seats, update meal preferences, and transfer to a human customer-service agent.
Alexa Plus on Amazon Echo speakers has a bug that makes it repeatedly sing 'lalala' for several minutes in the middle of conversations, without recognizing the behavior when users ask about it. The issue has been reported multiple times on Reddit in recent weeks across several Echo models, with the latest report three days ago. Amazon confirmed a bug affecting a 'small number of Alexa Plus users' and is working on a fix.
Napster co-founder Sean Parker is rebuilding Stability AI as an AI tools company for music professionals. Two years ago he participated in an $80 million rescue financing for the image-generation startup; he has now announced its new direction with CEO Prem Akkaraju.
AI music generator Suno added Speech, a feature that generates spoken audio together with matching background music and combines them into one track. Users enter an idea or text and describe the desired voice and musical style; the model generates the speech and sound.
Build your agents with the best Voice and Transcribe Streaming model in the world... now easily available to use on @vercel #1 for quality and speed... cheaper than any other hyperscaler... and a whopping 60% cheaper than Eleven Labs
Microsoft AI released MAI-Transcribe-2-Streaming, a real-time transcription model supporting 60 languages. Its first partial results arrive in just over 100 milliseconds. Microsoft says it ranks first for accuracy on Artificial Analysis; audio costs $0.54 per hour through the end of the year.
Suno launched Speech, a voice-generation feature now in public beta on its web and mobile apps. It generates a voice from a script or text prompt together with matching background music. The feature has Simple and Advanced modes; Advanced lets users customize the script and adjust the AI voice's gender, speaking style, and generation variety. Clips can be up to about 8 minutes, and background music can be switched off.
Andrej Karpathy shared several ways to ask LLMs to make their outputs easier to understand: write using the ASD-STE100 controlled-language standard or relax it to 80%; generate diagrams; output an interactive webpage in HTML; or make a custom explainer video, for example with voiceover using an ElevenLabs API key. He thinks that as models improve, more work will move up to supervision and understanding, and models will be able to create large custom software artifacts that previously were not worth making.
Tavus launched Griffin, which the company calls the first Human Interaction Model (HIM). In live video calls, it can receive and generate speech, facial expressions, tone of voice, gestures, and pauses at the same time.
Microsoft AI is launching MAI-Transcribe-2-Streaming, which Mustafa Suleyman calls the world's most accurate real-time transcription model. He claims it is 55% faster and 60% cheaper than ElevenLabs and invites developers to build agents on Microsoft's platform. In the quoted post, Artificial Analysis reports #1 rankings for final and first partial transcript accuracy on AA-WER Streaming, with final transcript WER of 2.5% at 0.13s after speech ends and streaming pricing of $0.54 per hour of audio.
Why it matters: The quoted benchmark comparison pairs transcription accuracy with latency and pricing, giving readers concrete tradeoffs to consider when choosing a streaming speech model for agents.
AI voice company ElevenLabs said employees can sell some vested shares at a $22 billion valuation, double its $11 billion valuation at a financing round in February. The $300 million tender offer was co-led by Wellington and T. Rowe Price. It is the company's second employee share sale, following a $100 million offer at a $6.6 billion valuation in September 2025.
Hugging Face released Open TTS Leaderboard, which uses objective metrics including WER/CER, RTFx, TTFA, and speaker similarity to evaluate open-source TTS models, reducing evaluation cycles from weeks of voting to hours.
Why it matters: Hugging Face replaces vote-based evaluation with objective metrics, showing how to compare open-source TTS models across languages and voice cloning.
On September 28, Alibaba's Qwen-Audio-3.1-TTS speech model ranked first globally on the Controlled Voice Arena leaderboard by Artificial Analysis with an Elo rating of 1,177.
An AWS tutorial shows how to deploy Qwen3-TTS on Amazon SageMaker AI using the vLLM-Omni DLC. A bidirectional SageMaker streaming connection sends text in and streams 24 kHz PCM audio chunks out, with a Gradio app for testing.
ElevenLabs released voice model Eleven v4, which better follows emotion, pauses, and sound-effect tags in scripts and keeps voices consistent in long productions. The new architecture also powers Turbo, a variant for real-time voice agents. In company tests, it begins speaking in about 150 milliseconds, compared with 262 milliseconds for Cartesia Sonic 3.6 and 814 milliseconds for OpenAI GPT-4o mini TTS.
Qwen3-TTS-12Hz-1.7B-Base from Alibaba Cloud's Qwen team is now available on Amazon SageMaker JumpStart and can be deployed to a fully managed real-time inference endpoint. It can clone a voice from just a few seconds of reference audio and its matching text, without retraining.
Google DeepMind has released Gemini 3.8 Live with Live Avatar, combining real-time video generation and voice to give Gemini's conversational model a near-real-time visual avatar, with accurate lip-sync, natural expressions, and smooth turn-taking.
Why it matters: Gemini 3.8 Live adds real-time video avatars, giving readers a view of how multimodal conversation is being applied in enterprise settings and where the feature is available.
AWS has released a WhisperX Deep Learning Container (DLC) that builds on OpenAI Whisper with batch inference, word-level timestamps from wav2vec2 forced alignment, and speaker diarization. It can be deployed to real-time or asynchronous Amazon SageMaker AI endpoints without building a custom image.
At Connect 2026, Meta put the Muse personal agent at the center of its hardware-plus-agent strategy, announcing voice and real-time video capabilities and several glasses devices, while previewing—but not releasing—a new frontier model.
🚀 Tencent Hy Translation just landed.Powered by Hy-MT2. 33 languages. 5 Chinese minority languages & dialects. Voice. Photo. Full offline — on-device, no network required. Already live in 12 countries and regions. Travel, drive, work or read abroad. Accurate. Natural. Always available. ⏬ ⏬⏬ https://apps.apple.com/us/app/tencent-hy-translate/id6801428074
@finkd just unveiled Muse Realtime Avatar, our real-time embodiment technology that turns Muse Realtime Voice into expressive, interactive avatars. Muse Realtime Avatar opens up a new way to interact in @Muse, starting with real-time conversations. 🧵👇 #MetaConnect
