September 28, 2026

ClaudeDevs

@ClaudeDevs

Claude can now help you build evaluations and hillclimb on them. In this article, we share guidance on eval design & skills that Claude Code can use to improve your applications. https://t.co/PgKFC2DWth

Sep 28, 2026, 08:54 PM UTC

♡ 9.1K likes↻ 606 reposts↳ 172 replies

OpenAI

@OpenAI

Get ready. https://t.co/bsf4j6vspM

Sep 28, 2026, 07:15 PM UTC

♡ 64K likes↻ 3.7K reposts↳ 3K replies

OpenAI Developers

@OpenAIDevs

We saved you a seat. See you tomorrow for the OpenAI DevDay keynote: https://t.co/NUSyPz99kp https://t.co/DlUDnZqQUC

Sep 28, 2026, 08:02 PM UTC

♡ 11K likes↻ 473 reposts↳ 378 replies

Anthropic

@AnthropicAI

Claude Sonnet 5.5 is now available:

Sep 28, 2026, 06:04 PM UTC

♡ 12K likes↻ 546 reposts↳ 410 replies

Simon Willison

@simonw

The most important thing about Sonnet 5.5 is that it's now the model that powers the free tier on https://t.co/f1IGIsU4Hc - so all of this stuff can be done by free users ChatGPT's free tier is still GPT-5.6 Luna, which is a lot less capable

Sep 28, 2026, 09:18 PM UTC

♡ 1.8K likes↻ 82 reposts↳ 187 replies

Cursor

@cursor_ai

Sonnet 5.5 is now available in Cursor! It’s a strong model, performing on par with Opus in many tasks. https://t.co/iaJ5z67FVu

Sep 28, 2026, 08:14 PM UTC

♡ 3.1K likes↻ 124 reposts↳ 145 replies

Cognition

@cognition

Today we're introducing updates that make Devin significantly cheaper. It's now: ▪️ 30–40% cheaper in Fusion and Normal ▪️ 15–20% cheaper in Ultra ▪️ Up to 70% cheaper in Devin Review All while improving capabilities – Devin Fusion now ranks first on FrontierCode 1.1 Extended. https://t.co/zj0yXe8FJ8

Sep 28, 2026, 06:03 PM UTC

♡ 1.1K likes↻ 96 reposts↳ 72 replies

Cognition

@cognition

Devin Mobile is now available in beta. Join the waitlist at the link below. https://t.co/OooQUYKmLT

Sep 28, 2026, 03:43 PM UTC

♡ 914 likes↻ 84 reposts↳ 102 replies

AI at Meta

@AIatMeta

In the past 6 months... - Muse Spark - Muse Image - Muse Video - Muse Spark 1.1 - Meta Model API - Muse Glimmer - Muse Spark 1.2 - Muse Code - Muse Spark 1.3 - Muse ...and we’re just getting started. 🚀 https://t.co/94e6uYPhZL

Sep 28, 2026, 02:30 PM UTC

♡ 2.8K likes↻ 146 reposts↳ 203 replies

OpenAI Developers

@OpenAIDevs

Meet the winners of The WebMCP Challenge. These 10 projects show what people and agents can build together when websites expose structured tools agents can use. 🧵 See the winning projects and the builders behind them: https://t.co/fynh6v6hCr

Sep 28, 2026, 05:00 PM UTC

♡ 1.6K likes↻ 127 reposts↳ 94 replies

Simon Willison

@simonw

I've published detailed notes and an annotated transcript to accompany the video of the keynote I gave at @WeAreDevs World Congress North America in San Jose on Friday - here's my rundown of everything that's happened with LLMs and agents in 2026 so far https://t.co/iENa7bKmdF

Sep 28, 2026, 05:51 PM UTC

♡ 709 likes↻ 88 reposts↳ 64 replies

Cognition

@cognition

Claude Sonnet 5.5 is now available in Devin Desktop and Devin CLI. On FrontierCode 1.1 Main, it scores 64.4%, a significant improvement from Sonnet 5 (56.2%), surpassing Fable 5.1 at extra high reasoning effort. https://t.co/DBlp4p0VB7

Sep 28, 2026, 08:30 PM UTC

♡ 302 likes↻ 20 reposts↳ 33 replies

Vercel

@vercel

If it's fast, it's on Vercel. We just made @MercedesAMGF1 even faster. Live at https://t.co/1AhAixBkw4 https://t.co/d4Htmru2L4

Sep 28, 2026, 07:25 PM UTC

♡ 377 likes↻ 32 reposts↳ 38 replies

Rive

@rive_app

Rive Editor 0.9 is live 🚀 Every feature planned for 1.0 is now in: keyboard and gamepad input, text fields, focus, accessibility, grid layouts, and free plan export. Next up is stability, so build something real and send feedback 👀 https://t.co/WcyZXBwSY2

Sep 28, 2026, 06:01 PM UTC

♡ 83 likes↻ 16 reposts↳ 7 replies

Google AI

@GoogleAI

https://t.co/TnNl6XmL15

Sep 28, 2026, 06:54 PM UTC

♡ 610 likes↻ 45 reposts↳ 40 replies

Visual Studio Code

@code

🆕 Claude Sonnet 5.5 is now generally available in @code with GitHub Copilot! Built for everyday coding tasks, it matched Sonnet 5 in early testing while using fewer steps, tokens, and tool calls. ⚡️ Give it a try today! 🚀

Sep 28, 2026, 06:53 PM UTC

♡ 284 likes↻ 18 reposts↳ 15 replies

Cognition

@cognition

https://t.co/rqcBypIUZr

Sep 28, 2026, 03:43 PM UTC

♡ 119 likes↻ 9 reposts↳ 10 replies

GitHub

@github

🆕 @AnthropicAI's Claude Sonnet 5.5 is now generally available in GitHub Copilot. Designed for well-scoped everyday work like building features and fixing bugs, Claude Sonnet 5.5 stood out for its efficiency in our early testing: It matched Sonnet 5 on coding tasks while using significantly fewer steps, tokens, and tool calls. Plus, it finished tasks noticeably faster. ⚡️ Try it out in the GitHub Copilot app, CLI, and @code. https://t.co/wZTQiZIEs7

Sep 28, 2026, 06:19 PM UTC

♡ 418 likes↻ 37 reposts↳ 44 replies

Simon Willison

@simonw

@WeAreDevs Here's how I define "Fable class models" - first Claude Fable 5, now Claude Opus 5.5 and GPT-Astra 6 and maybe GPT-5.6 Sol as well https://t.co/NHBcy0p2en

Sep 28, 2026, 05:53 PM UTC

♡ 74 likes↻ 3 reposts↳ 26 replies

Visual Studio Code

@code

🎬 How did VS Code make language support easier to build across editors? Explore the story behind language servers and the common protocol that helped shape the VS Code experience. 🍿 Watch The Story of VS Code: https://t.co/Q7cZIVisns https://t.co/S8mWSeYnBI

Sep 28, 2026, 02:15 PM UTC

♡ 244 likes↻ 26 reposts↳ 2 replies

Framer

@framer

Build more pages that actually feel like they belong on your site. Your /design-system skill keeps the same styles, components, typography, and layouts across every new page. https://t.co/i7vVOSiFw4

Sep 28, 2026, 08:00 PM UTC

♡ 67 likes↻ 7 reposts↳ 4 replies

LangChain

@LangChain

What separates a simple agent from one that can reliably work through complex, long-running tasks? It comes down to the harness around the model: how the agent plans, manages context, uses tools, and breaks work into smaller steps. In our upcoming virtual workshop, we’ll walk through these patterns and give you a practical look at how Deep Agents are built. Choose the session that works best for your region: 🌍 EMEA: https://t.co/f2IO2pnKKL 🌏 APAC + ANZ: https://t.co/8YHpHNNcCw

Sep 28, 2026, 03:02 PM UTC

♡ 86 likes↻ 11 reposts↳ 13 replies

LlamaIndex 🦙

@llama_index

Document parsing requires a lot of on-the-fly decision making. We explored the early approaches to Jev and Jev-like models on several documents tasks like orientation detection, language detection, routing, and more. https://t.co/U9urKwcORE

Sep 28, 2026, 09:00 PM UTC

♡ 31 likes↻ 6 reposts↳ 4 replies

Hugging Face

@huggingface

What would you like to see in Gradio? Drop your feature requests below 👇

Sep 28, 2026, 07:07 PM UTC

♡ 59 likes↻ 7 reposts↳ 54 replies

Cognition

@cognition

Model independence lets us call the best model for each part of Devin’s work. We continue refining the harness around them by batching more work per call, running more actions in parallel, and making the most of cached context. Read how the latest models and harness improvements deliver better price-performance: https://t.co/K0NSmT4mEM

Sep 28, 2026, 06:03 PM UTC

♡ 62 likes↻ 0 reposts↳ 5 replies

Simon Willison

@simonw

@WeAreDevs And my closing thought on how this stuff impacts our lives as software engineers, with a quote from cycling champion Greg LeMond https://t.co/67kIOqlVIY

Sep 28, 2026, 05:55 PM UTC

♡ 51 likes↻ 1 reposts↳ 8 replies

LlamaIndex 🦙

@llama_index

Parsing forms still trips up the latest frontier VLMs. A form isn't text on a page. It's a set of fields, grouped into sections, each tied to a specific box. That's why forms need purpose-built parsing, not a bigger general model: ✅️ Detect every field, not just the obvious ones ✅️ Keep the hierarchy of sections and fields ✅️ Tie every value to the exact box it came from ✅️ Understand handwriting and checkmarks Our latest blogpost breaks down where VLMs fail on real W-2s, 1040s, W-9s and scanned W-4s, and a custom cookbook for LlamaParse to handle them at a fraction of the cost 👇 https://t.co/KPTrjhvuby

Sep 28, 2026, 07:25 PM UTC

♡ 13 likes↻ 2 reposts↳ 8 replies

LangChain

@LangChain

🍽️ @wonder is building an agent that plans, orders and ensures delivery of all 21 of your weekly meals, while handling all customizations and customer service resolutions. With LangSmith, they swapped log debugging for LLM-as-judge scoring. A PM can file a bug and Claude Code fixes it from the trace. ⏯️ @ktech9999’s story

Sep 28, 2026, 07:00 PM UTC

♡ 33 likes↻ 3 reposts↳ 12 replies

OpenAI Developers

@OpenAIDevs

JupyterLite WebMCP brings an agent into your live notebook to edit cells, run code and review changes together. Built by Allison Coleman and Juan Mendoza. https://t.co/GFlSqjhdqF https://t.co/WB7YEkKU3X

Sep 28, 2026, 05:03 PM UTC

♡ 59 likes↻ 4 reposts↳ 8 replies

OpenAI Developers

@OpenAIDevs

Faraday by @thegreataxios lets an agent navigate 3D scans through WebMCP while the scan stays in the browser. Built for research and education. https://t.co/yTdVKmEwIb https://t.co/RlG4AhREyB

Sep 28, 2026, 05:03 PM UTC

♡ 51 likes↻ 2 reposts↳ 3 replies

OpenAI Developers

@OpenAIDevs

Aisle takes on the wedding seating puzzle with a WebMCP agent that respects guest relationships and pinned seats. Built by Eri Alipaj. https://t.co/dHlGPZVeh3 https://t.co/NlH64nlqQ2

Sep 28, 2026, 05:02 PM UTC

♡ 52 likes↻ 1 reposts↳ 4 replies

LangChain

@LangChain

Proud to be a launch partner on @nvidia's new Open Agent Safety Platform.

Sep 28, 2026, 04:56 PM UTC

♡ 36 likes↻ 1 reposts↳ 12 replies

LangChain

@LangChain

With LangSmith Engine v2, you can: 1️⃣ Proactively detect issues before they appear in production with Red Teaming 2️⃣ Quickly resolve agent issues with fixes validated in your environment 3️⃣ Detect harder-to-spot issues like inefficient agent work and metric trends https://t.co/Qb5rsGEmlC

Sep 28, 2026, 03:57 PM UTC

♡ 34 likes↻ 3 reposts↳ 5 replies

LangChain

@LangChain

This just in: @sydneyrunkle sat down with @jlowin and @zzstoatzz from @PrefectIO to talk about the new MCP spec + FastMCP, Jev, evals, and more! Give it a listen! https://t.co/MLnINAdPXL

Sep 28, 2026, 09:32 PM UTC

♡ 20 likes↻ 1 reposts↳ 6 replies

Cognition

@cognition

While the token prices for Sonnet 5.5 remain the same as Sonnet 5, Anthropic reports more than 30% faster output generation and up to 30% lower cost per task than Sonnet 5, with unchanged token pricing. https://t.co/YgOhlRTC21

Sep 28, 2026, 08:30 PM UTC

♡ 21 likes↻ 0 reposts↳ 3 replies

Rive

@rive_app

For a full list of updates you can see the announcement post here! https://t.co/Za51bTxEFt

Sep 28, 2026, 06:02 PM UTC

♡ 7 likes↻ 0 reposts↳ 1 replies

Simon Willison

@simonw

@WeAreDevs (The full talk also covers New Zealand’s Kākāpō parrot breeding season, probably the best news of 2026)

Sep 28, 2026, 05:56 PM UTC

♡ 21 likes↻ 1 reposts↳ 1 replies

School of Motion

@SchoolOfMotion

This week Greyscalegorilla tackles three everyday Cinema 4D headaches, Figma and Weave turn a trading-card design into a reusable workflow, and Andrew Price tests whether viral AI 3D demos survive contact with Blender. It's Motion Mondays! Watch the full episode now ➡️ https://t.co/tGSyl8MXMe

Sep 28, 2026, 04:01 PM UTC

♡ 12 likes↻ 0 reposts↳ 0 replies

GSAP

@greensock

💚 Site of the Day - Dmitrii Dark - Portfolio Dmitrii Dark - @DarkThePsy ⚙️ ScrollTrigger, ScrollSmoother 🛠️ Astro site → https://t.co/FSfKz7EIRt showcase → https://t.co/zGjV9MynPN https://t.co/GM4jPFAg3h

Sep 28, 2026, 02:43 PM UTC

♡ 15 likes↻ 0 reposts↳ 0 replies

LangChain

@LangChain

Usman Javaid, Group Chief AI Officer at @orange is headlining Interrupt, The Agent Conference by LangChain. Catch his talk + more on October 13th. https://t.co/GCNitijjvm https://t.co/hw4M2qZYPk

Sep 28, 2026, 01:02 PM UTC

♡ 17 likes↻ 3 reposts↳ 3 replies

Cognition

@cognition

Try it out today: https://t.co/y3humFLPXs

Sep 28, 2026, 08:30 PM UTC

♡ 6 likes↻ 2 reposts↳ 0 replies

Framer

@framer

Learn more about skills ↓ https://t.co/X8CJkBiqQr

Sep 28, 2026, 08:00 PM UTC

♡ 4 likes↻ 0 reposts↳ 0 replies

Rive

@rive_app

And some fixes can be seen here https://t.co/mrOj2qVNun

Sep 28, 2026, 06:02 PM UTC

♡ 4 likes↻ 0 reposts↳ 0 replies

Motionographer®

@motionographer

Dear Upstairs Neighbors | Connie He Sleepless nights. Noisy neighbors. An exhausted imagination running wild. In her six minute animated short, Pixar alum Connie He turns the frustration of insomnia into a colorful, increasingly unhinged journey. Behind it: 27 storyboard versions and a collaboration between filmmakers and Google DeepMind researchers, using generative models fine tuned on the team’s own concept art. For Motionographer® speaks with Connie about preserving comic timing, translating personal experience into production design, and helping artists and researchers understand what each needs to make a film. A conversation about craft, control, and what happens when a machine learns the visual details its artists hadn’t thought to explain. Read the full article: https://t.co/baBJTAnkns #DearUpstairsNeighbors #ConnieHe #Animation #MotionDesign #Filmmaking #mtngrphr #motionographer

Sep 28, 2026, 03:48 PM UTC

♡ 4 likes↻ 0 reposts↳ 0 replies

LangChain

@LangChain

Choose the session that works best for your region: 🌍 EMEA: https://t.co/RQEVRXIyy5 🌏 APAC + ANZ: https://t.co/t3KIXr75J8

Sep 28, 2026, 03:02 PM UTC

♡ 3 likes↻ 1 reposts↳ 1 replies