Launch
Show HN: Get your coding agent off ~2019 JavaScript, onto modern web primitives (opens github.com)
A Google Chrome team repo of guidance that steers coding agents toward modern browser primitives instead of 2019-era JavaScript.
Launch
Show HN: Every satellite that passed over any place this week, and its photos (opens watched.earth)
A browser tool that finds every imaging satellite that passed over a place in the last 7 days and pulls the real Sentinel-2, Landsat and NASA imagery.
Launch
Muse Ember: Our smallest, most efficient model (opens research.meta.ai)
Meta's entry-level model, described by SiliconANGLE as costing significantly less than the competition and the first trained using Muse Wonder.
News
DeepSeek says its open-source agent app Harness hit 1.54M DAUs on November 4, less than three months after launch (opens bloomberg.com)
DeepSeek says Harness reached 1.54 million daily users within three months of launch and processes about 92 trillion tokens a day, roughly 60 million per user.
DiscussionUnverified claim
Llama-modchip: 284B DeepSeek on 64GB RAM, no GPU (opens github.com)
Launch
Show HN: Coldframe, a honeypot that poses as a supermarket fridge controller (opens github.com)
A Show HN honeypot that poses as a supermarket refrigeration controller, built after recent attacks on industrial fridge installations.
Launch
Show HN: Agentic Identity Broker, OAuth Delegation for AI Agents (opens github.com)
An MIT-licensed OAuth2 broker from the Zalando Incubator that lets a user give an AI agent scoped access.
Launch
Show HN: Firenook – Firebase emulators in Rust: 40× faster writes, 10× less RAM (opens firenook.dev)
Hi HN, we have been developing with Firebase for a long time, and the only way Firebase developers can build apps without directly costing them is using the official Firebase Emulator Suite.
Launch
Thinking Machines Lab debuts Inkling 2 Small, a MoE model with 594B total and 17B active parameters, and plans to release weights next week (opens thinkingmachines.ai)
Artificial Analysis says it matches Gemini 3.8 Flash (high) on its index and would be the top open model outside China once weights ship.
Launch
Launch HN: Rivet (YC W23) – Scale and Orchestrate Agents on S3 (opens github.com)
Hey HN, this is Nathan and Nicholas from Rivet (https://rivet.dev) – a fast, high-density, and scalable orchestrator for agentic workloads using the actor model. Rather than running a heavyweight sandbox per agent, each agent gets a lightweight actor backed by S3.
Launch
Show HN: A decent text-to-image model trained for ~$20K (Apache 2.0 Weights) (opens linum.ai)
My brother and I run a 2-person lab that trains our own generative video models from scratch. The ultimate goal is a set of creative tools that make it easier for anyone to make animation. Video models today are too slow/expensive and lack sufficient controls. Creativity is an iterative process.
News
OpenAI details a pipeline that used an upcoming cyber model to validate and patch 948 local privilege escalation vulnerabilities in Linux for $1.2M in API costs (opens openai.com)
OpenAI details a pipeline that used an upcoming cyber model to validate and patch 948 local privilege escalation vulnerabilities in Linux for $1.2M in API costs — Security research that once required a separate, sustained investigation for each finding can increasingly be run as a systematic…
Launch
Show HN: Gauge – Open-Source CLI to recursively improve your docs (opens github.com)
Hi HN, I’m Evan from Gauge. We’re open-sourcing our CLI to recursively improve your docs for coding agents. The CLI includes a local execution backend, so you don't need an account to try it.
AnalysisOpinion
Qwen 3.8 27B on my laptop can interpret model weights, too (opens simonwillison.net)
Simon Willison follows up the paper "Sparks of Weight-Space Programming" and finds a local Qwen model can also read neural network weights.
Launch
Stringman, an $800 open-source room scale gripper robot (opens neufangled.com)
Neufangled Robotics' Stringman is an open-source gripper robot that works across a whole room, priced at $800.
Launch
Show HN: Kairos – Open-source risk and fraud decisioning platform (opens github.com)
Finic's Kairos is an open-source alternative to real-time fraud decisioning platforms such as Sift and Sardine.
Launch
Anthropic releases FlashProve, an open-source prototype that cryptographically verifies an AI response came from a given model, not one that a router swapped in (opens anthropic.com)
The prototype would let anyone check that a response really came from the model it claims to be from.
News
Revolut says it cut AI coding inference costs by ~90% in less than a year by making self-hosted GLM models its default over commercial tools like Claude Code (opens theregister.com)
Revolut says it cut AI coding inference costs by ~90% in less than a year by making self-hosted GLM models its default over commercial tools like Claude Code — Revolut says that in less than a year it's managed to slash its AI coding inference costs by roughly 90 percent by making self-hosted…
DiscussionUnverified claim
Official Statement: Qwen Models Will Stick to Open Source to the End! (opens reddit.com)
There were some rumours the last few days that Qwen4 will be the last open-weight series of Qwen, and future ones would be API only. The following statement contradicts the rumours and clearly states the stance that Qwen is taking regarding their future releases.
News
Memo: Linux distribution vendor SUSE is laying off ~10% of its employees, saying "automation and AI are fundamentally shifting how software companies operate" (Liam Proven/The Register) (opens theregister.com)
Memo: Linux distribution vendor SUSE is laying off ~10% of its employees, saying “automation and AI are fundamentally shifting how software companies operate” — One in ten employees at SUSE are facing the chop due to “automation and AI,” according to a memo dispatched by management …
Launch
Show HN: HackerBooks – HackerTube, but for books mentioned on HN (opens hackerbooks.org)
About 16.5k books mentioned by at least two HN commenters since 2007, each linked back to the comments; built with Claude after Karpathy's HackerTube.
Launch
Show HN: OmniDiff – Opensource, super fast, content aware Git difftool, in Rust (opens github.com)
OmniDiff diffs code in 25 languages plus config, images, PDFs, archives and fonts, and is written in Rust.
DiscussionUnverified claim
Anthropic's Mythos-class model allegedly leaked on Hugging Face 💀 (opens reddit.com)
A Hugging Face repo claims to hold an internal Anthropic Mythos-class training checkpoint; a follow-up post says the shards it downloaded are fake.
DiscussionUnverified claim
Memdoor: A team of 5 Qwen3.8-27B models beats Fable 5 on LiveCodeBench Hard (opens github.com)
The claim is that a team of five Qwen3.8-27B models orchestrated by Memdoor beats Fable 5 on LiveCodeBench Hard; it is the developer's own benchmark.
News
Making it easier for websites to express AI training preferences (opens openai.com)
Google, Microsoft, OpenAI and Anthropic will honour one shared robots.txt rule that keeps a site's content out of generative AI training.
Acquisition
Deepgram acquires Pipecat-maker Daily to bolster its voice AI infrastructure (opens techcrunch.com)
Deepgram is buying Daily, the company behind Pipecat, to deepen its real-time voice and video infrastructure.
Launch
Baseten’s joey-1 makes AI decisions for a penny per million tokens (opens thenextweb.com)
Baseten's joey-1 is a 2B open-weights decision model that returns an answer and a probability; Baseten says it costs $0.01 per million tokens and answers in about 20ms.
Acquisition
Sources: the US DOJ has inquired about the impact of Nvidia's Hugging Face deal on AI competition (opens reuters.com)
Sources: the US DOJ has inquired about the impact of Nvidia's Hugging Face deal on AI competition, as antitrust enforcers continue to probe Nvidia's conduct — The US Department of Justice has made inquiries about how Nvidia's (NVDA.O) $12.93 billion deal to buy open-source platform Hugging Face …