Trying Antigravity isn't an all-or-nothing choice with the new IDE extensions. Bring Gemini's multimodal reasoning into: • VS Code • JetBrains IDEs • Xcode • Zed • Visual Studio Extensions are a place to stay, or a place to start. antigravity.google/blog/antigra...

🤗 I've just released SetFit v1.2.0! SetFit trains text classifiers from a handful of labeled examples per class by fine-tuning a Sentence Transformer, no prompts or LLMs needed. v1.2 brings support for transformers v5, Sentence Transformers v6 & huggingface_hub v1. Thread 🧵

We just shipped Gemini 3.8 Flash! It's our third Flash model in six weeks. It really shines at long-horizon coding and autonomous agent tasks. To showcase the improved reasoning and tool-calling, check out this game written with a single looping prompt. Prompt and launch post are in the thread 👇

🚀 Zed v1.17 is out! Zed now previews CSV, TSV, PSV, and SSV files as interactive tables. You can sort and resize columns, then filter rows by selected values. Thanks HalavicH!

Pydantic AI's agent harness now has built-in support for Playwright! In a few lines of code, your agent can browse sites, click, run JS, and screenshot. To demo it, I built a QA agent that checks a website for issues (it already found a bunch on mine!): techcommunity.microsoft.com/blog/azurede...

Browser automation with Pydantic-AI + Playwright | Microsoft Community Hub

When we build agents, we often want to give them the ability to browse the web: open webpages, navigate from one page to the other, and read the content of a...

techcommunity.microsoft.com

⚠️ A few hours ago, a malicious crate was discovered on crates.​io which spread as a dependency of `arrayref` and some other crates, likely due to compromised credentials. The affected versions have been deleted. For details and how to see if you are impacted, see: blog.rust-lang.org/2026/08/20/s...

Supply chain attack on arrayref | Rust Blog

Empowering everyone to build reliable and efficient software.

blog.rust-lang.org

🚀 Zed v1.16 is out! Grouped change sections in the Git Panel are now collapsible. Thanks chirivelli!

🚨I've just released Sentence Transformers v6.0! MultiVectorEncoder joins the family: ColBERT-style late interaction models are now a first-class model type, for training, inference & interpretation, alongside dense, sparse & reranker models. Big thread 🧵

We will ship the first Polars 2.0 release candidate next week. Polars 2.0 will be great for casual Polars users as it will default to the streaming engine, meaning your code will be faster without doing anything. So the first release candidate next week. Give it a spin!

More actual lifesaving stuff from deepmind, which will probably not get too much coverage outside of nerd circles: deepmind.google/blog/weather...

AI model achieves breakthrough in forecasting cyclones

WeatherNext enables accurate cyclone forecasts that can give an extra day of warning. Now we are open sourcing the model.

deepmind.google

🤖 I've just released Sentence Transformers v5.7.0! A correctness & performance-focused release: gradient-cached losses got an overhaul (silently wrong gradients fixed, up to 3.9x faster training) & model.compile() now actually speeds up inference. + a long list of fixes. 🧵

👀 Sneak peek behind the scenes of Google Agent Skills! In my new article, I share how we build, test, and scale AI skills at Google: • Standardized layout • Remote MCP tools for IAM • Automated CI/CD checks • 2x2 evaluation matrix Full article link in replies 👇 #GoogleCloud #AIAgents #AgentSkills

Memory improvements in FastAPI 0.140.0 ⚡️ Around 50% less memory usage 🐬 The benchmark says 16x improvement, but that's an extreme case. It will depend on your app and usage of dependencies. Thanks to ipeluffo for the investigation! 🤓

Gemma 4 has been downloaded 300 million times since its launch in April! 🎉🚀💎 (And we got a mention in the earnings report 🥹) blog.google/company-news...

Q2 2026 earnings call: Remarks from our CEO

Read an edited transcript of Sundar Pichai’s remarks from the Q2 2026 Alphabet earnings call.

blog.google

#AI coding agents waste a lot of time and tokens repeating the exact same trial-and-error mistakes across sessions. To fix this, I created the 𝗢𝗽𝗲𝗻 𝗥𝗲𝗮𝘀𝗼𝗻𝗶𝗻𝗴 𝗙𝗼𝗿𝗺𝗮𝘁 (ORF) (a file-based specification that lets AI agents learn from past runs)

Transferring as much of the intuitions of building Olmo as I possibly can in the book format. The book is launching with an over 10 hour, full course with slidedecks, functional code for the training chapters, an example model completions library, and of course the free online web version.

Tencent just published R3-Skill, a two-stage retrieval stack purpose-built for a problem RAG-style retrievers weren't designed for: routing LLM agent skills (think Anthropic's SKILLmd format). Two 0.6B models, both Apache 2.0, one embedding model, and one reranker. 🧵

The outline panel gives you a compact overview of your search results. Scroll the full list of matches, click any one, and jump straight to it in the multibuffer. Read about it, and other hidden gems, in the latest post: zed.dev/blog/hidden-...

library-skills now supports uv workspaces and npm/Bun workspaces. ✨ Auto-repair and sync of skills. Improved support for .claude (additional to .agents), it will ask you interactively what to use. Plus several other bug fixes and improvements. library-skills.io

I’ll keep doing everything I can to make building things easier. 🤓 This should make it much simpler to build full-stack apps with FastAPI, for you and your agents. ✨ Psst... teach your agent to use it with: uvx library-skills

FastAPI@fastapi.tiangolo.com · 3mo ago

FastAPI can now serve your frontend app ✨ With support for client-side routing 😎 Great for React with TanStack Router, Astro static builds, Vite-based apps, etc. 🎉 FastAPI version 0.138.0 🔖 fastapi.tiangolo.com/tutorial/fro...

Starting with today's release (v1.7), Zed's native agent will automatically compact your conversation when your context hits a certain threshold.

I deployed an app from localhost → dev → production using an AI agent. No bash scripts. No gcloud commands. No YAML. Watch the full workflow 👇 youtu.be/cFSxmXQnU5Q

Antigravity: The AI Agent that Runs DevOps for Me

YouTube video by Cloud with Karl

youtu.be

🐛I've just released Sentence Transformers v5.6.0! A correctness- & robustness-focused release, headlined by a fix for a silent scoring bug in causal-LM rerankers (think Qwen3-Reranker), plus a batch of hard-negative mining & loss-correctness fixes. Thread with highlights 🧵

Just released ryl 0.18.1: the fast Rust YAML linter. New: a language server (ryl server) for inline lint, fix + format as you type in any LSP editor, a VS Code extension, and conda-forge availability. marketplace.visualstudio.com/items?itemNa... #rust #yaml #linter

🎉 Scikit-learn 1.9 released: ■Solid improvements to many existing estimators: faster, more stable, handling missing values, adding GPU support… ■Also, enhanced estimator displays in notebooks, ■And callbacks that enable progress bars or monitoring of convergence blog.scikit-learn.org/updates/rele...

Great software always took shape in conversation, not the commit. With agents, the conversation that generates the code is becoming the true source of our software. And Git can't keep up. So we built something that can. Meet DeltaDB: zed.dev/blog/introdu...

What would happen if we tried the diffusion generation on LLMs? We get Diffusion Gemma! 4x speed up! ⚡⚡⚡💎 blog.google/innovation-a...

DiffusionGemma: 4x faster text generation

An overview of DiffusionGemma, an exceptionally fast text generation model with up to 4x faster speeds.

blog.google

I wrote a new post for the Astral blog about how we’re building more vulnerability and malware defenses directly into uv: astral.sh/blog/uv-audit

Vulnerability and malware checks in uv

Find vulnerabilities in your Python dependencies with uv audit and prevent installation of known malware with uv's experimental malware detection.

astral.sh

🚀 Zed v1.5 is out! Mermaid diagrams render faster and more accurately now. 🧜‍♀️

Mellum started as a code completion model. Mellum2 goes much further – supporting both natural language and code. Now open source on @hf.co, Mellum2 is a 12B-parameter LLM for routing, RAG, and sub-agents, optimized for ultra-low-latency and high-performance inference. Learn more: jb.gg/awpava