Products from September 4, 2026

GPT-6 Astra is OpenAI's most capable model for complex reasoning, software engineering, computer use, science, and professional work. Async tool calling and mid-turn steering help it run multistep agent workflows. API: model id gpt-6-astra, $10/$50 per 1M tokens (short context). Rolling out first via Trusted Access/Daybreak, then Plus, Pro, Business, Enterprise, and the API in the coming days. First OpenAI model at the Critical cybersecurity capability threshold (advanced cyber access limited).

myAIcademy builds personalized AI training around your role, goals, and the tools you actually use. Learn through follow-along lessons, practice safely with simulations of real AI tools, and get step-by-step guidance from Aimy while you work. Generic courses and prompt libraries that quickly go stale, while myAIcademy continuously refreshes its content as AI changes, so your skills keep up too.

Compliance by TwelveLabs is a SaaS application that reviews your video library against rules your team writes, not ours. It ingests footage, applies your own compliance rule packs, and returns reviewer-ready findings with context, not just a timestamp and a label. Powered by TwelveLabs' Pegasus model, it explains why a moment may violate a rule so reviewers can accept, reject, or annotate in one queue.

Gemini 3.8 Flash and Gemini 3.8 Flash Cyber bring next-generation intelligence to agentic workflows and cybersecurity. Built for long-horizon coding, multi-step reasoning, autonomous tasks, and vulnerability detection, they deliver stronger performance at Flash speed and low cost.

Adding comments to your website shouldn’t require building a comment system from scratch. cmmnts gives you everything you need for discussions—threads, moderation, login, anonymous comments, Markdown, mentions, emoji, and GIFs. Drop it into your site and start collecting comments.

Omarchy is an omakase Linux desktop on Arch, Hyprland, and Quickshell. It ships a complete keyboard-first workstation with system-wide themes, packaged updates, and coding agents that can reshape the setup instead of making you assemble it yourself.

Our flagship AI weather forecasting model now includes real-time satellite data, hourly refreshes, higher resolution, precise precipitation forecasting, and clean energy variables. It’s now integrated across Search, Gemini, Maps, Google Maps Platform, and Cloud.

Snitch asks everyone in your Slack one question, who do you report to, and builds the org chart from the answers. Then it answers questions about it: reporting chains, team sizes, who owns what. No HR system, nothing for you to fill in.

TrackMCP shows who is using your MCP server, what they are trying to do, whether the work gets done, and where to improve. All from one line of code.

Review PRs from inside your running app: switch branches from an in-page widget and compare worktrees side by side. Loopback-only, zero dependencies.

Inline is a thread-based chat app for all work, with your team and agents.

A simple playground to write, run, and test JavaScript snippets, with an in-UI console. Built for quick experiments, debugging, learning, testing ideas, and working with JavaScript without needing a full development environment. This tool is offline by design, providing a robust JavaScript playground directly within Chrome.

Book AI coding agents on a real calendar. Clockwork runs them unattended in sandboxed git worktrees on your Mac, pauses for approval when something looks risky, and files a report with the actual dollar cost. Bring your own key or ride your existing subscription. Schedule recurring agent jobs on a real calendar, run them unattended in isolated sandboxes, and review honest reports.

RenoDX is an open-source framework for creating targeted graphics mods for DirectX games. It solves the problem of generic, one-size-fits-all post-processing by allowing mods to be built around individual games' rendering pipelines for precise upgrades like HDR output and shader replacement. It's for modders and players who want to enhance specific games with deep, persistent, and game-aware visual modifications.

Chalked is the reply layer for Mac. Open a supported conversation and it prepares what you would actually say from the visible thread, live calendar and sourced working context. Press Tab to insert, hold fn to change the intended outcome by voice, then review and send yourself. Accepted or edited communication can become inspectable evidence, so later replies start smarter. Voice is optional. Nothing auto-sends.

Miles is an open-source reinforcement learning framework designed to make training large language and vision models simple, efficient, and stable. It solves the engineering complexity of agentic RL for post-training tasks like SFT, RL, and distillation, enabling faster iteration and stable long runs. It's built for AI labs and enterprises working with frontier models, offering proven scale, extensive hardware support, and integrations with popular agentic environments.

This repository is a collection of public exploit Proofs-of-Concept (PoCs) and vulnerability research write-ups for unreported security flaws. It aims to lower the barrier to entry for security research by providing concrete, real-world examples for study and learning. It is intended for aspiring security researchers, bug bounty hunters, and students looking to understand vulnerability discovery and exploitation techniques in a practical, hands-on context.

OpenCode is an open-source AI coding agent that helps developers write code directly in their terminal, IDE, or desktop application. It solves the problem of fragmented AI coding assistance by integrating with any model provider, including local models, and works with existing subscriptions like GitHub Copilot and ChatGPT. It's built for privacy-first environments and is designed for developers who want a flexible, powerful, and open-source coding assistant.

Gemini 3.8 Flash and Gemini 3.8 Flash Cyber bring next-generation intelligence to agentic workflows and cybersecurity. Built for long-horizon coding, multi-step reasoning, autonomous tasks, and vulnerability detection, they deliver stronger performance at Flash speed and low cost.

Gemini 3.8 Flash is Google's most intelligent workhorse yet. Third Flash drop in 6 weeks, with big jumps from 3.7 on software engineering, agents, and multi-step reasoning. It sits with larger frontier models on DeepSWE v1.1 for long-horizon coding, leads Terminal-bench 2.1 at 89.4% and HLE-Verified at 54.9%, and ships at the same intro price as 3.7: $0.75/$3.75 per million tokens through Dec 31. Live today in the Gemini app (Pro/Ultra), AI Studio, Antigravity, Gemini Enterprise, and the API.