Things you might plug into production this week.
A technical comparison of Anthropic's Claude Sonnet 5, Sonnet 4.6, and Opus 4.8, focusing on agentic coding benchmarks, API pricing, and cost-performance metrics.
Anthropic Just Gave Its AI Coding Tool a Built-In Browser—Here’s Why Users Will Love It inc.com
OpenAI launches ChatGPT Work, introduces GPT-5.6 model family The Hindu
Nimbus builds production AI systems combining humans and AI end-to-end. From scoped pilot to production in 4 to 8 weeks.
Talk to Nimbus →OpenAI secures US regulatory green light for GPT-5.6 rollout, Axios report says MSN
Microsoft 365 Copilot to get GPT-5.6 default model IT Brief Australia
Stanford researchers released TRACE, an open-source, capability-targeted agentic training system designed to improve agent performance by turning recurrent failures into synthetic reinforcement learning environments.
Perplexity has introduced a migration guide from its Sonar chat completions API to its new Agent API.
Perplexity updated its "Brain" model, resulting in a 25% increase in answer correctness and 16% in recall, while simultaneously reducing task costs by 13% for workflows with prior context.
Skyfall AI released MORPHEUS, a persistent enterprise simulation benchmark designed to necessitate continual reinforcement learning in non-stationary environments.
A German research consortium has released Soofi S 30B-A3B, an open language model trained entirely on Deutsche Telekom's cloud infrastructure in Munich. The model uses an efficient hybrid architecture that activates…
Cursor builds AI agent 'Sand' to rival Anthropic's Claude Cowork TweakTown
Perplexity released new documentation guiding users to migrate from their previous Sonar chat completions API to the current Agent API.
SpaceXAI’s Grok 4.5 outperforms GPT-5.6-SOL, challenges AI leaders Crypto Briefing
Claude Code now has a built-in browser that lets the AI open, read, and interact with web pages directly inside the development environment. Write actions on external sites are screened by classifiers, and purchases or…
Theo – t3.gg: OpenAI’s GPT-5.6 Matches Fable at 1/38th the Cost, But It Won’t Stop Writing Code finance.biggo.com
OpenAI Introduces ChatGPT Work And Deprecates Atlas Browser Pulse 2.0
Anthropic Adds Built-In Web Browser to Claude Code Desktop App MLQ.ai
Claude Cowork's biggest use case is the mundane office work nobody wants to own, Anthropic says the-decoder.com
Perplexity launched "Perplexity Computer," a tool for go-to-market teams that uses orchestrated AI agents to perform end-to-end work such as competitive monitoring and generating sales prep kits.
vLLM v0.25.0 Release Notes Highlights This release features 558 commits from 232 contributors (64 new)! Model Runner V2 is now the default for all dense models (#44443). Building on quantized-model support from the…
OpenAI Integrated GPT-5.6 into Microsoft 365 Copilot incrypted
Ant Group's Robbyant unveiled LingBot-VA 2.0, a causal video-action model specifically designed for physical AI applications.
Meta Muse Spark 1.1 Earns 71 on Independent Coding Benchmark at One-Third Rival Cost Tech Times
Meta's Muse Spark 1.1 outperforms GLM-5.2 in coding and costs slightly less the-decoder.com
What's changed Auto mode is now available without `CLAUDE_CODE_ENABLE_AUTO_MODE` opt-in on Bedrock, Vertex AI, and Foundry; disable via `disableAutoMode` in settings Fixed the terminal freezing and keystrokes lagging…
Video | OpenAI Launches GPT-5.6 After Trump Administration Delay NDTV
OpenAI’s Turbulent Week: GPT-5.6 Launch, Executive Departure, Legal Setback and Product Shifts Mark Pivotal Stretch for the Company AI Insider
A big day for OpenAI.
Perplexity has launched "Deep Research" for its computer/desktop platform, designed to answer complex questions by generating comprehensive reports, dashboards, and decks.
OpenAI and Broadcom Reveal Custom Jalapeño Chip Optimized for LLM Inference theaisoftwarereport.com
According to OpenAI, GPT-5.6 Sol independently fine-tuned the smaller Luna model, triggered by a single "fairly under-specified prompt." In OpenAI's internal RSI benchmark for recursive self-improvement, Sol scores…
OpenAI Turns ChatGPT Into a Work Agent The Neuron
OpenAI rolls out GPT-Live for real-time voice AI globally MSN
OpenAI launches ChatGPT Work to bring AI coding power to professionals AnewZ
OpenAI debuts GPT-5.6 and ChatGPT Work to bring AI agents into the workplace The Indian Express
Patch Changes 356918c: feat(provider/openai): add GPT-5.6 reasoning and prompt cache controls
OpenAI launches GPT-Live for real-time AI voice conversations MSN
Anthropic launches AI platform for scientists, plans own drug development
Perplexity introduced 'Computer for Counsel', an AI assistant designed to orchestrate agents across multiple frontier models, files, and tools to handle complex legal tasks.
Meta Launches New A.I. Model as Global Technology Race Heats Up The New York Times
Meituan has released LongCat-2.0, a 1.6 trillion parameter MoE language model featuring a native 1 million token context window.
LlamaIndex introduced 'legal-kb', an agentic retrieval system using Index v2 with specialized retrieval and text-processing tools.
Claude Science: What Anthropic’s New Research Workbench Actually Does for Oncology Oncodaily
Advances in Claude’s autonomous AI capabilities MSN
OpenAI launches GPT-5.3 Instant: Smoother, smarter ChatGPT experience The Eastleigh Voice
US lifts curbs on Anthropic’s Fable, Mythos AI models DD India
Meta says its next AI model has caught up with OpenAI’s GPT-5.5 Storyboard18
NVIDIA introduced HORIZON, a hands-free agent framework designed for hardware design that treats the process as repository-level code evolution.