OpenAI set to buy Ona
OpenAI announced plans to acquire Ona, whose secure, persistent cloud environments will strengthen its Codex coding agent. The deal addresses a key limitation of agentic coding—ephemeral sessions that…
Tool Reviews
Hands-on reviews of AI apps, assistants and image, video and voice tools: pros, cons, pricing and who each one is for.
OpenAI and Broadcom introduce a custom chip designed for LLM inference
OpenAI and Broadcom have introduced Jalapeño, a custom AI chip purpose-built for large language model (LLM) inference. Unlike general-purpose processors, it aims to deliver faster responses, better en…
A first look at GPT-5.6 Sol: the next-gen model
OpenAI has teased GPT-5.6 Sol, a next-generation model focused on three demanding domains: software development, scientific research, and cybersecurity. It ships alongside OpenAI's most advanced safet…
The Rise of ChatGPT Adoption: How It Has Grown
New OpenAI Signals data shows ChatGPT adoption is deepening worldwide. Users are shifting from curiosity to daily habit, exploring more features like writing, coding, and learning support, and adoptio…
GPT-5.6: Cutting-edge intelligence that grows alongside your ambitions
GPT-5.6 represents cutting-edge AI intelligence designed to grow alongside your ambitions. It delivers better quality from every token, maximizes value for each dollar spent, and provides greater capa…
Education Innovation Highlights AI Testing and Accessibility
Two Education Innovation items group together as a single AI-and-education update. The first item is titled “Testing LLM
Microsoft 365 Copilot now favors GPT-5.6 as its top model choice
Microsoft has updated Microsoft 365 Copilot to run on GPT-5.6 as its top model, improving reasoning, instruction-following, and output quality across Word, Excel, PowerPoint, Copilot Chat, and the new…
AI education and science entries focus on accessibility, skills, research, and infrastructure
A closely timed set of entries from the same site groups together short Education Innovation and General Science items t
A report card for the era of artificial intelligence
OpenAI CFO Sarah Friar proposes a practical scorecard for evaluating AI investments: useful work completed, cost per successful task, dependability, and return on compute. This framework helps busines…
Unveiling OpenAI Presence
OpenAI Presence is an enterprise-focused AI agent platform that enables organizations to deploy trusted voice and chat agents for both customer service and internal workflows. Supporting spoken and wr…
How GPT-5.6 brings together cutting-edge intelligence and cutting-edge efficiency
GPT-5.6's headline achievement isn't flashier capabilities but efficiency. OpenAI improved performance across three layers: the core models, inference (the biggest recurring cost), and agentic workflo…
Pushing the boundary of cost-efficiency: Introducing GPT-5.6
OpenAI's GPT-5.6 update significantly lowers pricing for its Luna and Terra models, designed for high-volume business tasks. Driven by optimized inference and right-sized architecture, the cost reduct…
Enhancing GPT‑5.6 Sol in ChatGPT and bringing GPT-5.6 Luna access to free users
ChatGPT's GPT-5.6 family gets a major upgrade. GPT-5.6 Sol, the stronger model, delivers improved accuracy and consistency for complex tasks like writing, research, and coding. Meanwhile, free users g…
Exclusive seating options are on the way for ChatGPT Business
OpenAI introduced Premium seats for ChatGPT Business, offering roughly 5x the usage of standard seats and removing the five-hour usage limit entirely. Admins can mix Premium and standard seats across …
Unveiling @huggingface/kernels: Over 200 WebGPU Kernels for On-Device AI
Hugging Face launched @huggingface/kernels, a JavaScript library that loads 207 optimized WebGPU kernels from the Hub for fast in-browser AI inference. Each kernel ships as a versioned package with ma…
BenchMIRT: What Do LLM Benchmarks Really Measure?
BenchMIRT is a new auditing method for LLM benchmarks that applies multidimensional Item Response Theory (MIRT) at the individual prompt level to uncover which underlying capabilities—such as safety o…
A first look at Ultrafast mode: GPT-5.6 Sol delivering speeds up to 14 times faster
OpenAI has launched Ultrafast, a preview tier running GPT-5.6 Sol on Cerebras wafer-scale hardware, delivering up to 14x faster generation and 750 tokens per second. The article explains why latency i…
The Guardian's Pane
AI has become a weapon for attackers and a powerful ally for defenders. This article explores how AI is reshaping cybersecurity, how OpenAI is strengthening its defenses by preventing model misuse, pr…
Jalapeño's initial results reveal record-setting speed and efficiency in AI inference
OpenAI's custom Jalapeño chip is purpose-built for AI inference, delivering record-setting speed, higher throughput, lower latency, and improved power efficiency. The article explains what inference i…
Excited about iOS 27's impressive AI capabilities and revamped Siri? Be prepared for usage caps
Apple has released a new support page explaining how to manage the AI-powered features arriving with iOS 27, including the revamped Siri. The page documents where users can turn features on or off, co…