The morning edition
Stay tuned.
Keep building.
A short daily dispatch from the front lines of AI. Three things worth knowing, before you get to work.
Five minutes or shorter.
One clear picture.The daily show
for AI builders
Today, in three stories.
The written edition. Your next move included.
01 / GPT-6.2 Astra and Sol
OpenAI Introduces GPT-6.2 Astra and Sol AI Models
OpenAI has introduced GPT-6.2 Astra for complex professional work and GPT-6.2 Sol for efficient everyday tasks.
Confirmed by vendoropenai.com02 / Gemini Decisions API
Google Launches Gemini Decisions API for Fast AI Decisions
Google launched the Gemini Decisions API for fast, probabilistic choices across text, images, audio, video, and PDFs using the Gemini Decision 1 model.
Confirmed by vendorblog.google03 / Hy4 flash
Tencent Launches Hy4 flash Model
Tencent released its Hy4 flash model, which offers multimodal input support, a 1M-token context window, and high decoding speeds at a low cost.
Reportedtencent.com
You're caught up. Three useful things. Back to building.
Have more time? Explore the wireSafe to ignore Nothing was left out on purpose. More stories are on the front page.
Join the bottom for $1, or take any rank for 10% more than the total paid for it, at least $1 more.
Payments are not open yet. Checkout is in test mode and takes no money.
Sponsored placements keep the brief free. Put your product here
The wire.
Go deeper in the things you build with.
- 01
OpenAI Introduces GPT-6.2 Astra and Sol AI Models
OpenAI has introduced GPT-6.2 Astra for complex professional work and GPT-6.2 Sol for efficient everyday tasks.
- 02
Microsoft's agent-framework PyPI package compromised
Versions 1.26.1 and 1.26.2 of the agent-framework PyPI package were compromised to steal Azure, GitHub Actions, and other CI/CD credentials.
- 03
Google Launches Gemini Decisions API for Fast AI Decisions
Google launched the Gemini Decisions API for fast, probabilistic choices across text, images, audio, video, and PDFs using the Gemini Decision 1 model.
- 04
OpenAI launches ChatGPT Wallet for secure agent payments
OpenAI introduces ChatGPT Wallet, a new way for agents to make secure payments without sharing actual card details.
- 05
Tencent Launches Hy4 flash Model
Tencent released its Hy4 flash model, which offers multimodal input support, a 1M-token context window, and high decoding speeds at a low cost.
- 06
Qwen4-Max Now Available on Vercel AI Gateway
Alibaba's Qwen4-Max model is now available for use on Vercel's AI Gateway.
- 07
Mervin Praison Reviews Edge0-145B Local AI Model
Mervin Praison reviewed Edge0-145B, a free local AI from Edge0, detailing its performance, resource needs, and setup steps.
- 08
Claude Launches Whiteboards, Sheets, and Data
Claude adds new tools for team collaboration, spreadsheets, and data work inside its product.
- 09
Self-hosted Claude Managed Agents enters private beta
Developers can now run Claude Managed Agents, including the control plane, in their own cloud or on-premise environment.
- 10
Answer.AI launches SolveIt AI collaboration tool
Answer.AI released SolveIt, a tool for people to learn and work with AI, code, and documents together.
- 11
Langfuse Launches AI Gateway in Public Beta
Langfuse introduced its AI Gateway in public beta, offering a central entry point to manage and trace all AI usage.
- 12
Coding Agent Study Finds 64% Cost on Failed Runs
A large-scale evaluation of AI coding agent harnesses found that most inference costs came from runs that did not succeed.
- 13
- 14
- 15
- 16
- 17
- 18
- 19
- 20
- 21
Gemini Robotics 2.5 boosts robot skills and safety
Google DeepMind's Gemini Robotics 2.5 model enhances robot skills, speed, and safety for real-world tasks.
- 22
- 23
- 24
Google Search Upgrades with Gemini 4 Hydrogen
Google Search now uses the Gemini 4 Hydrogen model globally for more intuitive, visual, and precise answers.
- 25
OpenAI Shows Codex Working With Video Editing and Agent Workflows
OpenAI's Dominik Kundel and Charlie Guo showed how Codex can edit video and connect agent workflows at CODE Summit 2026.
- 26
NVIDIA Vera Rubin NVL72 leads MLPerf agentic inference performance
NVIDIA Vera Rubin NVL72 shows high agentic inference performance in MLPerf Endpoints v1.0 results.
- 27
Together AI’s Together Compiler uses AI agents to beat megakernels
Together AI reports that its Together Compiler uses AI agents to rewrite and fuse GPU kernels in model forward passes, and that this approach outperforms its previous hand-written megakernels on B200.
- 28
Simon Willison Highlights Meta AI’s Muse Ember Pricing and Efficiency Data
In a link post, Simon Willison points to Meta AI’s description of the Muse Ember model and shares pricing and Artificial Analysis efficiency results.
- 29
- 30