-
Town raised $55M to build an AI assistant that learns how you work
## What it is Town is a personal AI assistant that adapts to you instead of the other way around — your voice, your people, your patterns. It works across email, calendar, docs, and messages, drafting replies, handling your scheduling, and building routines for the work you do over and over. The more it watches… Continue reading
-
Hermes is a self-improving AI agent you run on your own server
## What it is Hermes is an autonomous AI agent that lives on your own server and gets more capable the longer it runs — it remembers what it learns across sessions instead of starting cold every time. You reach it from a terminal, from messaging apps, or now through Hermes WebUI, a browser and… Continue reading
-
Devin Desktop puts every local and cloud agent on one Kanban board
## What it is Cognition just turned Windsurf into Devin Desktop, an agent-neutral IDE built for running fleets of AI agents at once. The centerpiece is the Agent Command Center: a Kanban board of every agent you’re running — local and cloud — sorted by status (in progress, blocked, ready for review). One agent refactors… Continue reading
-
Microsoft Intelligent Terminal turns the command line into an agent orchestrator
## From shell to agent console Microsoft’s Intelligent Terminal, shown at Build 2026, is Windows Terminal rebuilt for the agent era. You invoke agents in plain English, watch their output in real time, and chain them together like Unix pipes — except the “commands” are autonomous agents. A built-in agent called ShellMgr turns vague requests… Continue reading
-
Windows Agent Studio lets you build Windows agents with drag-and-drop or TypeScript
## What it is Microsoft launched Windows Agent Studio in preview at Build 2026 — a tool for building custom agents that run natively on Windows. You get two paths: a visual drag-and-drop canvas for non-coders, or TypeScript for developers who want full control. Agents can reach local files, apps, and sensors when you grant… Continue reading
-
A Netflix engineer’s Headroom cuts LLM token bills up to 95% — and it’s open source
A Netflix senior engineer just open-sourced the tool you wish you’d written. Headroom (LLM context compression) jumped 1,000+ GitHub stars in a single day, and the pitch is brutally simple: most of the tokens you’re paying for are junk. What it actually does Headroom sits as a transparent proxy between your app and any of… Continue reading
-
NVIDIA Physical AI Agent Tools & Skills (NemoClaw + OpenShell) turn robots and AVs into agent tasks
NVIDIA just open-sourced a big batch of agent skills for Physical AI — and to be clear, this is a software stack, not a chip. The pitch: take the messy pipelines behind robots, self-driving cars, and factory digital twins, and make them executable by any coding agent. What it actually is The core is the… Continue reading
-
—
title: “Microsoft MAI-Thinking-1 outscores Claude Sonnet 4.6 in blind evals — trained without OpenAI’s data” date: 2026-06-03 tags: [model, microsoft, reasoning, api] Microsoft shipped its first fully self-built reasoning model at Build 2026, and the signal is hard to miss. MAI-Thinking-1 hits 97% on AIME 2025 and 94.5% on AIME 2026. In blind human evaluations… Continue reading
-
Microsoft MAI-Code-1 Flash is live in GitHub Copilot — 60% fewer tokens than comparable coding models
Microsoft shipped its first self-trained coding model at Build 2026. MAI-Code-1 Flash is a 5B-parameter model trained entirely inside GitHub Copilot’s own production tool harness — real developer workflows, not synthetic benchmarks. That’s the unusual part: the training environment is the deployment environment. The numbers hold up. 85.8% on Microsoft’s adversarial coding benchmark, ~51% on… Continue reading
-
Tiny-vLLM Rebuilds the Inference Engine in C++ and CUDA So You Can Read Every Kernel
Tiny-vLLM is a fully functional LLM inference engine in C++ and CUDA — written deliberately small, with every kernel and every line of math walked through in a free companion course. It’s the “younger and smaller sibling” of vLLM, built for understanding rather than deployment. ## What’s actually in there The implementation is not a… Continue reading
Top AI Product
Every day, hundreds of new AI tools launch across Product Hunt, Hacker News, and GitHub. We dig through the noise so you don't have to — surfacing only the ones worth your attention with honest, no-fluff reviews. Explore our latest picks, deep dives, and curated collections to find your next favorite AI tool.
Top AI Product 