Tag
Google's new AI agent Gemini Spark can operate browsers and make phone calls to complete purchases, while their Enterprise Agent Platform is now generally available with adoption by major companies.
The author built a local realtime voice stack using Parakeet STT, Qwen 2.5 7B, and Qwen3-TTS, integrated with Ollama.
A developer shows how to run a 28M parameter LLM locally on an Amazon Echo Dot 2 using llama.cpp, achieving around 4 tokens/s generation and ~2.3s latency with prompt caching, enabling simple offline voice commands.
Khoi Nguyen introduces Zinley, an AI with its own phone number and inbox that acts as a private personal assistant, capable of answering calls, sending email summaries, and working across call, email, desktop, web, and iMessage.
Apple's revamped Siri AI has finally arrived in iOS 27 beta, delivering on its promise of a personal, context-aware assistant, but it feels anticlimactic given advances in AI elsewhere.
SpeakoFlow is an open-source local voice assistant for desktop, allowing users to interact with their computer via voice without cloud dependency.
Microsoft is testing a new native real-time voice model, MAI Realtime, in early access on its MAI Playground. The full-duplex system supports multiple languages, low latency, and configurable turn-taking, positioning it as a competitor to OpenAI's GPT Live and Sesame.
The author shares Parlor v2, a fully local GPT-Live clone running on an M3 Pro, built as a cascade system after failed attempts to fine-tune Gemma 4 12B for full-duplex behavior.
CYBER is a JARVIS-style voice-controlled desktop AI assistant with an Iron Man HUD, PC control, webcam vision, and remote Telegram access; v3 rebuilds the interface with real-data visualizations.
Friend, the AI wearable companion, returns with a new built-in speaker for voice interaction and a higher price of $249, though its purpose beyond conversation remains unclear.
claw-code is an open-source project written in Rust that uses AI to automatically manage museum exhibits, offers audio guides and interactive Q&A, requires no manual maintenance, and the project itself is written and updated autonomously by AI.
Crodo AI is a voice-first AI assistant for macOS, showcased as a new product launch on Product Hunt.
Alex Finn shares how using ChatGPT Voice on a hike through the redwoods enabled him to get more work done than at his desk, leading Sam Altman to agree and express a desire for a new kind of computer.
Suggests using ChatGPT voice as a chief of staff to orchestrate tasks, delegate, and check status via voice commands.
Amazon's Alexa Plus gets an AI update enabling deeper smart home integrations with brands like Bosch, Delta, and Ecovacs, plus new partnerships via the Model Context Protocol.
The article reveals the secret origins of Amazon's Alexa and Echo, from Jeff Bezos's 2011 whiteboard sketch to the technical challenges of far-field speech recognition and the team's race to market after Apple's Siri launch.
AnovaX is a local-first desktop voice assistant that uses an LLM planner and multi-agent orchestrator to execute tasks on the user's computer. It features a safety layer, adaptive recovery, and a phone-friendly remote interface.
The Verge reports that the new Siri AI upgrade in watchOS 27 beta makes the Apple Watch feel like a true wrist computer by providing a consistent AI experience across phone and watch, enabling more complex tasks directly from the wrist.
A guide and open-source project to turn a Big Mouth Billy Bass into a real-time voice assistant using Strands Agents and Amazon Nova 2 Sonic on a Raspberry Pi 5.
Hugging Face has open-sourced a modular speech-to-speech tool, which separates the four layers (VAD, speech recognition, LLM, and speech synthesis) and supports replacement, allowing users to flexibly swap components or run locally without rewriting the entire pipeline.