-
Cognition SWE-1.7 hits 42.3% on FrontierCode — a few points behind Opus 4.8, at a fraction of the cost
The Devin team shipped SWE-1.7 on July 8, their strongest coding model yet. It’s not a chatbot and not an app — it’s the agentic model that drives Devin’s autonomous software engineering: give it a task, it reads the repo, edits files, runs the terminal, and closes the loop on long-horizon work. The number that… Continue reading
-
Meta Model API (Muse Spark 1.1) undercuts OpenAI and Anthropic at $1.25/$4.25 per million tokens
For years Meta kept its best models to itself and gave the weaker ones away. On July 9 that ended. The company launched its first paid product — the Meta Model API — and put Muse Spark 1.1 behind it. Zuckerberg called the price “roughly 25% of what Anthropic and OpenAI charge.” Bloomberg, CNBC, and… Continue reading
-
Anthropic put Claude Cowork on your phone — and 90% of users aren’t coding
Anthropic shipped Claude Cowork to web, iOS, and Android on July 7, opening in beta to Max subscribers first. Cowork is an autonomous agent: you hand it a task, it runs, and now it doesn’t need your laptop open to keep going. What actually changed The desktop app was leash-length — close the lid, work… Continue reading
-
OpenAI ChatGPT Work takes direct aim at Claude Cowork with an always-online agent
OpenAI shipped ChatGPT Work on July 9, and it’s not a chatbot. You hand it an outcome, not a prompt. It pulls context from your connected apps and files, breaks the goal into steps, and grinds through them for hours, then hands back a finished doc, spreadsheet, deck, report, or even a website. What it… Continue reading
-
NanoKVM-Go puts AI agents outside the OS — where BIOS and lock screens live
Every computer-use agent has the same blind spot: it lives inside the operating system. It can’t touch a BIOS menu, a Windows lock screen, a frozen machine, or a fresh OS install. NanoKVM-Go is Sipeed’s fix, and it’s hardware, not software. A watch-sized KVM that acts as physical hands It’s a KVM dongle. One USB-C… Continue reading
-
Meta Muse Spark 1.1 opens to developers — and Meta charges for AI for the first time, $1.25/$4.25 per million tokens
Meta’s Superintelligence Labs shipped Muse Spark 1.1 on July 9. It’s Meta’s move into the AI coding and agentic arena, aimed squarely at Anthropic and OpenAI. And for the first time ever, Meta is charging money for a model. What it actually is A multimodal reasoning model built for agent work. It plans multi-step tasks,… Continue reading
-
OpenAI: Separating signal from noise in coding evaluations — the lab just called its own benchmarks unreliable
Everyone ships a coding model with a shiny SWE-bench number attached. OpenAI just published a technical post arguing most of those numbers are noise. It hit 216 points on Hacker News, and the reason is obvious: it’s OpenAI throwing cold water on the exact scoreboard the whole industry uses to declare victory. What it actually… Continue reading
-
Microsoft Flint (a visualization language for AI agents) compiles one spec into Vega-Lite, ECharts, and Chart.js
Ask an LLM to draw a chart and you usually get a wall of Vega-Lite config, half of it broken — flipped axes, dead scales, labels stacked on top of each other. Charts are the last mile where agents still look dumb. Microsoft Research’s answer is Flint, an open-source intermediate language that sits between the… Continue reading
-
Meta cuts 8,000 jobs and admits its AI agents stalled for four months
This isn’t a product launch. It’s a confession. Meta laid off roughly 8,000 people — about 10% of its workforce — and then Zuckerberg told staff at a July 2 town hall that AI agent development “hasn’t really accelerated in the way we expected” over the past four months. The reality behind the pivot Meta… Continue reading
-
Willow Frontier Pro & Frontier Mini: a YC startup made cloud dictation free to break Wispr Flow’s paywall
Dictation used to be a bad trade: slow local models that ran on-device, or fast cloud ones that charged you monthly. Willow, a YC-backed voice company, shipped two speech-to-text models at once to break that deal. Two models, one aimed squarely at Wispr Flow Frontier Mini is the headline: cloud-based dictation that’s free and unlimited,… 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 