Build a Coding Agent That Knows Your Code—For Less Than a Weekend Project Cost

The Allen Institute for AI (AI2) has introduced a new open-source project aimed at making powerful coding agents more accessible and affordable. Called Open Coding Agents, the release centers on a method to train these systems for a fraction of previously reported costs.

Your Research Paper, But Without the 12-Tool Juggling Act: OpenAI Launches Prism

OpenAI has introduced a new, free tool designed specifically for scientific work called Prism. It is described as an AI-native workspace for scientists to write and collaborate on research, powered by their GPT-5.2 model.

The Soloist That Can Conduct an Orchestra: Meet Kimi K2.5

A new and powerful open-source model has been introduced, named Kimi K2.5. Building on its predecessor, it was trained on a massive dataset of approximately 15 trillion visual and text tokens

Your Spreadsheet Just Woke Up: Claude is Now Living Inside Excel.

Spreadsheets are no longer silent grids of numbers. Claude AI has moved in

Clawdbot: An Open-Source, Self-Hosted AI Assistant for Your Favorite Messaging Apps

Clawdbot has been released, offering a framework for running a personal AI assistant directly on a user's own hardware.

Microsoft Research Introduces Rho-Alpha, an AI Model for Robotic Manipulation

Microsoft Research has announced Rho-alpha (ρα), a new artificial intelligence model designed to translate natural language commands into control signals for robotic systems.

Paranoid for driver No More: Tesla Cuts the Safety Cord on Robotaxis.

Tesla has initiated its first robotaxi rides without a human safety monitor in the vehicle, according to an announcement by CEO Elon Musk

Researchers Develop Detachable Robotic Hand That Can Operate Independently

Researchers have presented a new design for a robotic hand that can detach from its arm to move independently, grasp multiple objects, and then return and reattach

Google DeepMind just Introduced D4RT: A New AI Model for 4D Scene Reconstruction and Tracking

A new artificial intelligence model named D4RT (Dynamic 4D Reconstruction and Tracking) has been introduced by Google

Anthropic Upgrades Task Management in Claude Code from Todos to Tasks

Anthropic has announced an upgrade to the task management system within its Claude Code developer tool, evolving from simple Todos to a more robust Tasks primitive.

RunwayML Launches Gen-4.5 Image to Video, Emphasizing Storytelling and Control

RunwayML has announced the launch of an Image to Video capability for its latest AI model, Gen-4.5.

Lightricks Partners with ElevenLabs to Launch Audio-Driven Video Generation

Lightricks has announced a new feature for its LTX-2 model: the ability to generate video directly from an audio file

Anthropic Publishes New 'Constitution' to Guide Claude AI's Values and Behavior

Anthropic has released a detailed new document called \"Claude's Constitution,\" intended to serve as the foundational guide for the values, ethics, and intended behavior of its Claude AI models.

Z.ai Announces GLM-4.7: A New Model That Shows Gains in Coding Benchmarks and Agent Performance

ai has announced the release of GLM-4.7, the latest iteration of its large language model.

xAI Releases Open-Source Code for X's 'For You' Feed Recommendation Algorithm

xAI has published the core source code for the recommendation system that powers the “For You” feed on X (formerly Twitter)

OpenAI Announces US Launch of Low-Cost ChatGPT Go and Plans to Test Advertising

OpenAI has announced two key updates aimed at expanding access to its AI tools: the broad launch of its affordable ChatGPT Go subscription tier and plans to begin testing advertisements within the platform.

Vercel Labs Releases 'Agent Skills' to Enhance AI Coding Assistants

Vercel Labs has released an open-source project called Agent Skills, a collection of packaged instructions and scripts designed

Wing and Walmart to Expand Drone Delivery Service to 150 Additional Stores Nationwide

Drone delivery company Wing and retail giant Walmart have announced a significant expansion of their partnership, planning to scale their residential drone delivery service to 150 new Walmart stores across the United States over the next year.

Replit Launches Mobile App Builder That Turns Descriptions Into Published Apps

Replit has launched a new feature, Mobile Apps on Replit, designed to let users create and publish mobile applications through natural language descriptions, bypassing traditional native development setups.

FLUX.2 [klein] AI Model Family Launches, Targeting Real-Time Image Generation and Editing

The FLUX.2 [klein] family of AI models has been released, designed specifically for high-speed, interactive image generation and editing.

Manus Announces Data Partnership with Digital Intelligence Leader Similarweb

Manus has announced a new data partnership with Similarweb

Google Launches ‘Personal Intelligence’ Beta, Allowing Gemini AI to Access Personal Gmail, Photos & More

Google has introduced a new beta feature called “Personal Intelligence” for its Gemini AI, enabling a deep, opt-in connection to a user’s personal Google Apps.

Google Announces MedGemma 1.5 and New Medical Speech AI, Citing Accelerated Healthcare Adoption

Google has released a significant update to its suite of open-source medical AI models, highlighting the healthcare industry's rapid adoption of artificial intelligence

OpenAI Brings On Torch to Build a Consolidated 'Medical Memory' for AI Health Tools

OpenAI has announced that the health technology startup Torch is joining the company

New AI Architecture 'Engram' Aims to Boost Model Efficiency with Dedicated Knowledge Retrieval

Researchers from DeepSeek-AI and Peking University have introduced a new architectural module for large language models (LLMs) called Engram, detailed in a technical paper published in January 2026.

Sakana Researchers Introduced DroPE, a Method to Extend AI Model Context by Removing Positional Embeddings

Sakana AI and the University of Oxford has introduced a new method, dubbed DroPE

Boston Dynamics Launches Commercial Atlas Robot and Won Best of CES 2026 Award.

Boston Dynamics has announced the commercial launch of its Atlas humanoid robot, transitioning it from a research platform to an enterprise product designed for industrial automation. The company also announced that Atlas won the best of CES 2026 award.

Anthropic Releases Official Code Simplifier Plugin for Claude

Anthropic Introduces Code Simplifier Plugin for Claude

Gmail Announces AI Features for Summaries, Writing Help, and Inbox Prioritization

Google has announced a suite of new AI-powered features for Gmail, calling it the platform's entry into \"the Gemini era.\"

Lightricks Open-Sources LTX-2, Delivering Production-Ready Audio-Video AI

Lightricks has released the full LTX-2 model as open source. The multimodal model, capable of generating synchronized audio and video, is now publicly available with complete weights and a suite of development tools.

OpenAI Introduces ChatGPT Health: A Separate Hub for Wellness and Medical Information

The team behind ChatGPT has announced a new, separate area within its platform focused specifically on health and wellness

NVIDIA Introduces Cache-Aware Streaming ASR for High-Concurrency Voice Agents

NVIDIA has released Nemotron Speech ASR, an open automatic speech recognition model designed to address scalability and latency challenges in real-time voice AI applications.

Boston Dynamics and Google DeepMind Announce Partnership for AI-Powered Humanoid Robots

Boston Dynamics and Google DeepMind have formed a new partnership aimed at integrating advanced artificial intelligence into humanoid robotics.

NVIDIA Unveils Open-Source AI Framework for Reasoning-Based Autonomous Vehicles

NVIDIA announced the Alpamayo family, a suite of open AI models, simulation tools, and datasets designed to accelerate the development of autonomous vehicles (AVs) capable of human-like reasoning.

DeepSeek New AI Training Framework Amid Global Competition

DeepSeek has published a research paper outlining a more efficient approach to developing artificial intelligence, highlighting the industry's push to advance despite constraints on accessing advanced semiconductors.

Tesla Driver Reports Coast-to-Coast Trip Using Full Self-Driving Without Manual Intervention

A Tesla driver has reported completing a roughly 2,700-mile journey from Los Angeles, California to Myrtle Beach, South Carolina using the vehicle's Full Self-Driving (FSD) system without manually taking control at any point.

PrimeBOT Unveils Compact, Open-Source Humanoid Robot for Personal Use

The Q1, has been introduced by PrimeBOT—a joint venture between AgiBOT and Shangwei New Materials.

Nvidia Introduced NitroGen: A Foundation Model Trained on 40,000 Hours of Gameplay

A collaborative research team from NVIDIA, Stanford, Caltech, and other institutions has introduced NitroGen, a vision-action foundation model designed to create generalist AI agents for video games.

Introducing Pickle 1: AR Glasses Framed as a Proactive AI Companion

Pickle has introduced the Pickle 1, a pair of augmented reality (AR) glasses positioned not just as a wearable display, but as what the company terms a \"Soul Computer.\" The device is presented as a proactive AI companion designed to integrate into daily life.

From Spectacle to Substance: The Necessary Pivot of AI in 2026

Microsoft Chairman and CEO Satya Nadella, outlined a 2026 outlook for artificial intelligence, marks a notable pivot in the industry’s discourse.

Manus Joins Meta: A Strategic Move to Embed AI Agents Across Platforms

Meta announced that Manus is joining the company. This strategic move aims to integrate Manus's advanced autonomous AI agent technology across Meta's consumer and business products.

Qwen3-TTS models introduce two powerful capabilities—Voice Design and Voice Cloning

How AI is Learning to Speak with Feeling and Character

Forget the Shiny AI Tools. First, Do This.

A practical, step-by-step framework to identify where AI can drive real ROI—without the hype.

Beyond the Hype Cycle: The Google AI Announcements That Actually Mattered in 2025

Google has laid out its own scorecard: a sprawling, 60-announcement chronicle of a year spent embedding artificial intelligence into every facet of its ecosystem.

When AI Can't Speak UI: Google's A2UI Teaches Agents to Show, Not Just Tell

A2UI (Agent-to-User Interface), aims to bridge. It’s not an app you’ll download, but a foundational standard—a kind of \"HTML for agents\"—that could reshape how we interact with AI.

From Prompt to Play: YouTube's Gemini-Powered Game Builder Enters the Arena

YouTube, the titan of video sharing, is stepping onto a new playground with the closed beta launch of Playables Builder, a prototype web app that lets users generate simple games using AI—no coding required.

Beyond Playlists: ChatGPT Wraps Your Year in Code and Conversation

OpenAI is rolling out \"Your Year with ChatGPT,\" a feature that aims to summarize your 2025 in prompts and pixels.

Beyond Seeing and Hearing: Meta's New AI Model Unifies Sensory Perception in a Single Brain

Researchers from Meta’s Superintelligence Labs are unveiling a significant step toward closing this gap. They’ve introduced Perception Encoder Audiovisual (PE-AV), a new family of AI models designed to natively understand and link audio, video, and text within a single, unified system.

Breaking the App Barrier: ElevenLabs Brings AI Agents to the World's WhatsApp Chats

The company is expanding its omnichannel conversational AI platform to support WhatsApp, bringing real-time voice and chat agents to the world's most popular messaging service.

From Chat to Action: Google Shrinks the AI Agent to Fit in Your Pocket

FunctionGemma, a new, specialized variant of its compact Gemma 3 270M model, engineered expressly for reliable function calling at the edge.