Why do you have 5 AI subscriptions? 👀 Get Pieces & use them all at the same time.

Tired of overpaying for premium AI? Switch to Pieces and access all your favorite models along with 9+ months of long-term memory. Save time, money, and context. 🧠

Introducing "Pieces of Memory". A brand new AI webinar series by builders, for builders. 🛠️ Bi-weekly sessions on: ➡️ Building long-term memory for AI ➡️ Productivity hacks that actually scale ➡️ Latest LLM/agent architectures First episode on July 10th at 9am. Sign up below.

OpenAI solved the intelligence problem. Pieces solves the memory problem. Context from Slack. Context from your standups. Context even when apps crash. This is what it looks like when your copilot actually remembers your workflow.

Why pick one... when you can run all of them WITH long-term memory? 👀

Are AI agents about to reshape how we build software? In a recent piece, we break down why this shift isn’t just about faster code generation or better autocomplete. The teams that figure out how to work with agents (not just use them) will be the ones that win. Read the full article here 👇

Stop copy-pasting old solutions. With Pieces + Cursor integration, just ask 'How did we fix this last time?' and get instant access to your past solutions. Your entire development history, searchable through natural language.

Clicked through 20 Chrome tabs just to find the blog post you skimmed yesterday? We've been there. That's exactly why we built Pieces. It's like having a photographic memory for every useful thing you've ever saved. But better.

In 2017, a small group at Google built the transformer architecture. Today? It’s quietly running almost everything. ChatGPT? Transformer. ElevenLabs? Transformer. Suno? Transformer. The building block behind it all.

Your AI shouldn’t start from zero every time. Pieces Long-Term Memory gives it: - 9+ months of context - Your PR history - Slack threads - Docs and notes Answers grounded in your actual work.

We recently updated Pieces MCP Server. Better retrieval. Smarter memory generation. And now you can manually generate memories right inside Cursor, GitHub Copilot, and more. Have you tried it yet?

No code? No problem. At the Pieces AI Summit, Andrey Gavrilenko showed how you can break into data science with just English and an LLM - no Python required. Check out the full recap 👇

Ever wonder how you actually solve long-term memory for AI? In a recent interview, Ellie Zubrowski sat down with Principal Research Scientist Antreas Antoniou to chat about nano models, transformers, local AI, and privacy. Check out the full interview here & let us know what you think! 👇

We've all been there - explaining the same project context over and over to different AI tools. That ends now. Pieces MCP Server creates a bridge between your development history and AI assistants, all running locally on your machine.

The 10x devs of the future will all be using AI. Will you be one of them?

Pieces OS 12.0.0 is live, introducing sign-in for all users. This isn’t just about security. It’s a foundational step toward a more reliable, consistent, and personalized experience across the Pieces ecosystem. Smarter memory, better agent experiences, and tighter integrations are all on the way.

Think of it as your personal coding time machine. Every snippet, solution, and breakthrough you've saved in Pieces is now available right in your Cursor chat. Your past work, powering your future solutions.

If your work involves juggling 10+ projects at once, Pieces basically becomes your second brain. Not just for code - docs, slides, random links, meetings... all of it.

Our CEO as an action figure 🤭

SEO isn't dead, but it's evolving faster than ever. While old tactics still technically work, they're becoming less effective as AI reshapes how people search for information. Time to adapt or get left behind.

Be honest, is this you? 👀

Your LLMs just got a memory upgrade.

Is this the future? 🤔

Join us for the Pieces Community Sync - today at 1pm ET. (in about 1 hour!) This is a good chance to connect with the team, ask questions, share feedback, and see what others are building with Pieces. Make sure to set a reminder 👇

You ever sit in a dev meeting and think, “How am I even here?” What triggers your impostor syndrome the most? 🤔

If remembering things isn’t your strong suit... outsource it to Pieces 🧠

Going live in a few hours 🎥 Episode 1 of The Missing Pieces drops today @ 12pm ET. Ellie Zubrowski is chatting with Hanna Stechenko, our Head of SEO, about her path into devtools and how her work helps devs discover Pieces. 🔗👇

🚨 New series alert 🚨 Meet the people building Pieces 👋 The Missing Pieces premieres tomorrow - June 4 @ 12pm ET. First up: Ellie Zubrowski sits down with Hanna Stechenko, our Head of SEO, to talk about her career path and how devs end up finding Pieces in the first place.

Not all AI models need to be massive. As Sam Jones (Chief AI Officer at Pieces) put it: “I think 90% of people working in software, if you tell them this does the same job, it’s cheaper, it’s faster, and you don’t have to ping an API… I think everyone wants that.”

We love seeing this kind of feedback from our Discord community 🫶

Happy Monday devs! What are you shipping this week?

Rewind AI isn’t what it used to be… and honestly, it never really worked for developers. If you’ve been thinking about switching, this post walks through the best memory-first tools in 2025 - from privacy-safe screen recorders to code-aware long-term memory agents.

“I’ve vibe coded a little too close to the sun before.” - @blackgirlbytes.bsky.social‬ If you’ve been using agents, Claude, Goose, or MCP - this talk is worth your time. Rizel Scarlett breaks down what vibe coding actually means, how it works, and what can happen when things get out of hand.👇

In LTM‑2.5, Workstream Activity got a serious glow-up 👀 Temporal groupings, improved navigation, related Copilot chats - a whole new hierarchy in the UI to keep you oriented and in flow. More context, less confusion. Have you tried the new Workstream Activity view? We'd love your feedback!

Can you use AI in the middle of nowhere? 👀

Everyone’s been focused on building smarter models, but recently the focus has shifted. The real advantage now lies in owning memory, context, and the ecosystems where agents operate. Google, Microsoft, OpenAI, Meta - they’re already moving in that direction. Check out this article to learn more👇

Vibecoding: The End of Computer Science? ft. @scott.hanselman.com‬

With @getpieces.bsky.social LTM-2.5 just dropping, listening to Tsavo Knott, Brian Lambert, Antreas Antoniou and @elliezub.bsky.social talk about it 🔥

Join us for the Pieces Community Sync - today at 2pm ET. This is a good chance to connect with the team, ask questions, share feedback, and see what others are building with Pieces. Make sure to set a reminder 👇

LTM‑2.5 is one of our biggest updates yet. And today, we’re talking about it live. Join Tsavo Knott, Brian Lambert, Antreas Antoniou, and Ellie Zubrowski to cover: - What’s new in Workstream Activity - How nano models made everything faster - Improvements to Pieces MCP Server Set your reminder👇

Anyone else feel like this after a long weekend? Yeah, same. With Pieces installed, getting back into work is way easier. It remembers what you were doing - files, links, chats, even bugs. You can even ask Pieces Copilot what’s next on your todo list. 😎

LTM‑2.5 landed last week, and tomorrow we’re talking all about it. Join Tsavo Knott, Brian Lambert, and Antreas Antoniou live with Ellie Zubrowski to cover: - What’s new in Workstream Activity - How nano models made everything faster - Key takeaways from #MSBuild Set your reminder 👇

You might forget what you worked on last Tuesday. But Pieces doesn’t. It has the notes, links, and code - ready when you are. What’s something Pieces remembered that surprised you?

Happy Monday devs! What are you shipping (or trying to ship) this week?

🚨 New series alert 🚨 The Missing Pieces kicks off June 4 @ 12pm ET. Every other week, we’re talking with the people behind Pieces - how they got here, what they actually do, and everything you won’t find on a resume. Set a reminder!👇

Pieces LTM‑2.5 just dropped, and we’ve got a lot to unpack. From nano model upgrades to major workflow memory improvements, this isn’t just a patch. It’s a full-on leap. 🗓️ Wednesday, May 28 at 12pm ET Bring your questions. Set your reminder. See you there 👇

Join us for the Pieces Community Sync - Thursdays at 2pm ET. This is a good chance to connect with the team, ask questions, share feedback, and see what others are building with Pieces. Make sure to set a reminder 👇

Rewind: 14GB/month, $19, no work summary, Mac only. Pieces: 0.22GB/month, free tier, AI rollups every 20min, Mac/Windows/Linux. Oh - and we actually remember what you were working on. Try Pieces. It's like your brain, but searchable 🧠

With LTM-2.5, you’ll actually remember what you were working on. 😉