Ho ho. you can now try @MistralAI 's new Magistral Small on @huggingface with @FeatherlessAI . *rubs hand*. Hmm. gonna have some fun. 🤓

RT @FeatherlessAI: ⭐The Year of Inference is here. Featherless is now an official inference provider on Hugging Face, unlocking 6,700+ LLM….

RT @darin_ver: Today, Featherless became the largest LLM inference provider on Hugging Face. 6,700+ models. Zero GPU config. One platform.….

RT @FeatherlessAI: What a month to start with @aiDotEngineer . 🏆 Featherless Action R1—the world’s most reliable agent that runs daily task….

I wonder when AI companies will start selling your data to advertisers and embedding ads in their chat messages for users on the free tier. Or is it already happening and I'm just not aware yet. 🤔.

RT @picocreator: A few folks have asked, how does the action-R1 agent we built at @FeatherlessAI in collaboration with @ui_licious look lik….

RT @kimmonismus: Things are not looking good for career starters. Artificial intelligence is increasingly threatening the traditional entry….

11/. ⚡ This is the future of RPA and UI testing automation:. Smart. Adaptive. Reliable. And we’re building it. Stay tuned! 😉. 👉 Want early access? Join the waitlist:.

10/. We're working on building TAMI : Auto-Pilot mode, an agent for exploratory testing to help QA teams:. 🖥️ Identify features.🧐 Discover edge cases.🧪 Design test scenarios. Because exploratory testing is still a completely manual, tedious, and boring process today. And it.

9/ . That's why at @ui_licious, we're working on developing agentic AIs that can:. 👀 Observe the UI.🧠 Understand features and components.🛠️ Use the UI intelligently. to perform simple to complex tasks - like a human user. 🧠. It can generate, update, and fix automation.

8/ . Now, imagine the future AI-generated, personalized UIs. I think it's not far. How then? Existing automation tools are too rigid to work reliably. 💡 We need robust and adaptable tools. 👉 The solution is agentic AI that can see, understand, and adapt - like human users.

7/ . You know fixing the same broken automation script over and over again is?. Insanity. 🤪. It is common for me to meet teams that have completely abandoned RPA projects and UI test automation efforts because maintenance overhead became too much work.

6/ . Companies spend billions on RPA and UI automation tools that are just glorified macros. that produce rigid scripts and require constant maintenance to keep up to date with the changing UI. Even the best teams I know have 10-20% of their scripts disabled or ignored,.

5/. Why does this matter? . Why do we care so much about building reliable AI agents? . Imagine a tool that performs a task correctly 75% of the time. Would you use it?. Ask any CTO or tech lead. I'd drop a tool that requires babysitting. But that's the hidden pain in the RPA.

4/ . Featherless-Action-R1 leverages @ui_licious low-code web automation framework and smart targeting engine to achieve stunning accuracy. 🎯🎯🎯. On some sites, it's hitting 100% successful task completion - with repeatability. Where as other agents are capping out at 80%.

3/. Yes, this is an AI that can actually do your boring chores. 😀. Not play Pokemon (Oye! Leave my video games alone!). 😑. What Siri was promised to do! 🙃.

2/. What's REAL?. Realistic Evaluations for Agents Leaderboard. It's a benchmark that tests AI agents in performing 100+ real world tasks such as:. 🍔 Ordering food.📅 Creating calendar events.✈️ Booking flights & hotels.📩 Sorting mail. across replicas of popular sites -.

1/. Featherless-Action-R1 scored 65% success on the REAL benchmarks. where most agents plateau at ~40%. And the next best? Anthropic-Computer-Use - at 43%. That's not a lead - it's a massive leap ahead. 😤😤😤.

🧵👇 THIS is the future of Action AI, RPA, and UI test automation. I'm really proud (grinning really), to celebrate the breakthrough that our collaborators - @picocreator and the @FeatherlessAI team has made in action AI. Their agent — Featherless-Action-R1 — just achieved.

What's a hard task? 🤔. A pretty ordinary thing that many people are doing everyday: . Searching for a place to buy/rent and filling a contact form to request for a tour at the earliest possible date.