Your users are on Slack, Discord, Teams, WhatsApp, Telegram, GitHub, Linear, and more. Your agents should be too. Chat SDK lets your agents run on every platform from a single codebase. Watch the announcement ↓

Every chat platform has its own event model, threading system, and streaming quirks. We felt this pain internally when we challenged every team to build agents to multiply their output, and the agents were easier than the chat plumbing. So we built Chat SDK to remove that bottleneck. Get started directly or via your coding agents with: ▲ ~/ 𝚗𝚙𝚖 𝚒 𝚌𝚑𝚊𝚝 ▲ ~/ 𝚗𝚙𝚡 𝚜𝚔𝚒𝚕𝚕𝚜 𝚊𝚍𝚍 𝚟𝚎𝚛𝚌𝚎𝚕/𝚌𝚑𝚊𝚝 Read more ↓

Just used v0 to build a slack bot in under 30 minutes Last year I could barely spell agent Incredible tooling and endless use cases

One of the best Vercel projects, absolute banger

I was like ahh I build this cool shit slack app that allows me to work on any github repo with any stack from my slack. And then I watch this vid to find out Fernando has done it as a side gig :( And that Vercel just has solved my problem that I did not yet knew I have 😭

This is actually a solid move. Meeting users where they already are instead of forcing them into a new tool just makes sense. If agents can run across all these platforms without breaking the experience, that’s a big step forward. The “single codebase” part is what really stands out if it genuinely reduces the effort to ship and maintain across platforms, that’s huge for teams. Simple idea, but high impact if executed well.

Exciting integration possibilities there

Everyone's excited about omnichannel agents. But the real bottleneck isn't platform coverage—it's context coherence. An agent that remembers your user's Slack conversation when they switch to Discord is worth 10x more than one that posts everywhere with amnesia.

this is the "write once, deploy everywhere" moment for AI agents. building separate integrations for Slack, Discord, Telegram was a nightmare — especially for indie devs who can't afford to maintain 5 different bot codebases. one SDK that handles all the platform quirks? shut up and take my money.

How straightforward is it to build a custom adapter for a platform you’re building yourself? Thinking about teams that want to unify agent behavior across both external channels and their own product’s interface.

Distribution across platforms from one codebase is the right primitive. But Chat SDK solves agent-to-human on every surface. The missing half: agent-to-agent. Your Slack agent discovering a specialist agent on Linear it was never configured for. Cross-platform presence is step one. Cross-agent discovery is the unlock.

This is really exciting!!!

write once hallucinate everywhere

claude code announced telegram / discord channels 2 hours after this announcement, coincidence? 🤔

just implemented it in @useprio for

Just tested Composer 2 on my iOS project — the multi-file edits are noticeably smarter. It refactored my SwiftUI views across 4 files without breaking a single @State binding. Curious how it handles large codebases with 50+ files. Anyone stress-tested it yet?

this is so cool!

i'm quite awed. its possible that chat sdk will have more impact on agents than ai sdk had on workflows.

This is the right direction. Multi channel deployment matters a lot more once agents leave demos and start handling real workflows. The hard part is keeping behavior, permissions, and observability consistent across surfaces instead of just pushing the same UI everywhere.

when we adding imessage?

@montonenico 🔥

This is major. Thanks, Vercel.

More 🔥🏗️✨✨

We love Vercel

@cramforce This is good. Do we know if this is going to be fedramped?

@cramforce would love to work on something together here

That's amazing!

@fernandorojo This is so cool. I bet it comes with a skill 😉

@HarounPapi did u see this?

▲

vercel is just crazy

A unified backend can make cross-platform consistency feel seamless, which often drives higher engagement. Balancing feature parity with platform conventions can be tricky at scale.

Thats is veeeeery good for custom agent

the single codebase for any platform is what makes this real. dispatching tool calls across different platform SDKs is the ugly part of multi-platform agents. this abstracts that at the right level