Today at ZapConnect, we debut Next Gen Zaps. One of the most important product moments in @Zapier's history. With Next Gen Zaps, customers can deploy agent-built Zaps in seconds. Simply describe your problem to your agent and it will build you a Zap that runs automatically, even when your laptop is closed. Use it from Claude, ChatGPT, etc. to migrate existing Zaps, create new ones, or bring external automations onto Zapier. It's one line of text away. My two favorite parts: 1. Hardening: Zaps will run the deterministic parts with code not AI. That means more reliability and lower cost. 2. Healing: When a Zap breaks, it falls back to agentic mode and can figure out how to recover without human intervention. The early data shows how much we still have to learn. But we think it's good enough to share now. One early access customer told us, "This is what Zaps were always supposed to be: easy." Now in beta. Let us know what you think: zapier.com/s/ngz-join-wait-l…

Build a Zap right from Claude, ChatGPT, Muse, Grok Bot, and more. That's Next Gen Zaps, and it just had its biggest week yet. Describe it once. @Zapier builds and runs it, even when your laptop is closed. It picks the apps, wires up the steps, and deploys the Zap for you. Then it keeps running on its own. 3 big differences: 1. Predictable: the steps that don't need AI run as code 2. Easier than ever to build: you describe it in plain language instead of building step by step 3. It fixes itself: if a run breaks, an agent works out a fix and tests it (the bigger changes ask you first) Open beta is coming soon. In the meantime, request early access here: zapier.com/workflow/early-ac…

Zapier is a launch partner for the Personal Agent Protocol: A new open standard from @SierraPlatform and @Meta for how personal AI agents work with companies. A growing share of your customers will reach you through an agent, and they'll never open your website. The Personal Agent Protocol gives agents a direct way in. Here's what it means for your company: 1. Every agent says who it's acting for, so you know which customer is on the other end. 2. Your customer decides what the agent can do: look things up, or take action 3. You decide where agents can work: your website, your APIs, or your own agent. The companies that are easy for agents to work with? Will win that customer. The rest get skipped without ever knowing they were in the running. @Zapier has spent 15 years getting software to work with other software. This is the evolution of that job. This is a work in progress, and a lot will change. But the rules for how agents work with companies are being written right now, and we want to help write them.

If you're using Grok Bot, add the Zapier plugin! Instant access to 9,000+ more apps. If you've tried it, share what you're building.

Zapier MCP is growing 4x faster than Zapier did at a similar scale. People want to build from anywhere and everywhere. For most of @Zapier's 15 years, you opened the Zap editor, picked a trigger, and wired up each step yourself. But now people start wherever they use AI: Claude, ChatGPT, Muse, etc. They describe what they need, and the AI reaches their apps through Zapier MCP. Every new AI tool becomes a front door to Zapier. In Muse's launch week, more people added Zapier MCP there than to ChatGPT. All of that growth is still what our team calls "fingertip work." Someone asks their AI to do one thing in their apps, Zapier MCP does it, and then it waits for the next request. The next frontier is automations that run on their own... That's 𝗡𝗲𝘅𝘁 𝗚𝗲𝗻 𝗭𝗮𝗽𝘀, in early access now. Describe the job once, wherever you build, and it deploys a Zap that keeps running after you close the chat (or the laptop). It can even fix itself if something goes wrong. We're bringing Zapier to a whole new world of builders. If you want to last 15 years in tech, you have to reinvent yourself. To try Next Gen Zaps early, join the waitlist here: zapier.com/s/ngz-join-wait-l…

Your Muse agent, after Zapier. Same agent, connected to 9,000+ more apps. @Muse already has over 3 million daily users. But it doesn't have native connectors for tools like HubSpot, Airtable, LinkedIn, X, etc. Connect @Zapier MCP? Now Muse can reach all of those, plus 30,000+ actions across every app on Zapier. A few prompts to try once Zapier is connected: 1. "Add everyone who registered for last week's webinar to HubSpot as contacts." 2. "Turn this blog post into a LinkedIn post and an X post. Show me both and publish once I approve." 3. "Every Friday, recap what changed in our pipeline." 4. "Find every customer feature request from this week, log each one in Airtable, and build a chart." Admins keep their rules, too. They pick which apps and actions each team can use, and every action is tied to the person who ran it. It's a safer way to let your team try new agents. @Zapier works in Muse, Claude, ChatGPT, Grok Bot, and anywhere you build with AI.

Our team trusts Next Gen Zaps with work that can't go wrong. Like 200+ speaker reminder emails on the day of ZapConnect. Our social team asked for use cases in @Zapier's Slack, and replies from the team flooded in. A lot of them were core parts of people's jobs. That’s some of the best product feedback you can get. A few that jumped out: One Next Gen Zap oversaw our biggest event of the year, ZapConnect, and sent 200+ speaker reminders, each in its delivery window. Ryan described it to Claude and had it built in a few hours. Another runs every time we close a deal. It pulls the Gong calls and product usage, then writes up why we won. Emily's writes her Friday status update and posts it to her working group. 7 Fridays straight without her touching it. James' Next Gen Zap drafts his daily team handoff from everything he worked on and sends it to him as a pre-filled form. When he hits approve, it goes to the team in Slack. Ben's keeps our engineering docs up to date across our internal repositories, so they don't go stale. Rob's reads every reply to our early-access emails each weekday, flags builders worth a customer story, and drafts outreach to each one. Zach's balances the scheduling load across our on-call sales rotations. All of these are Next Gen Zaps built with a prompt, in a tool like Claude or ChatGPT. Even our engineer Alex, who's been building "artisanal Zaps" 😂 by hand for 7 years, wasn't sure he'd like handing control to an agent. But now he builds them faster than he could by hand, and the ones he shipped haven't needed any babysitting. Our people know how to keep AI governed when the stakes are high: by building a human step into the Zap itself. None of the speaker reminders went out until Ryan approved them. Rob's agent drafts the outreach, and he reviews it before anything reaches a customer. James gets a pre-filled daily handoff in his DMs, edits it, and hits submit himself. It’s up to you. Low-risk work runs on its own, and anything headed to a speaker or a customer waits for a person. Next Gen Zaps are in early access on Pro, Team, and Enterprise. Join the waitlist here: zapier.com/s/ngz-join-wait-l…

Claude Haiku 5.5 is out today. 10x cheaper, and a huge improvement over 4.5 @ClaudeAI's Haiku 4.5 costs $1 per million input tokens and $5 per million output. Haiku 5.5 is $0.10 and $0.50. It's also the first Haiku with effort levels, so you pick how much it thinks (and how many tokens it spends) on each request. Note: the 10x price slash only holds while a request stays under 100k input tokens. Past that, rates actually jump 5x. That's still half of what 4.5 costs. But anything that stuffs a long history or a ton of info into a single request could cross that line. Big upgrade over Haiku 4.5. See all models @Zapier's scored here: zapier.com/benchmarks

Self-healing is one of my favorite parts of Next Gen Zaps. Now if a Zap breaks, it can fix itself. Approve each fix, or let the agent handle it. How it works: build your automation in whatever AI tool you like (Claude, ChatGPT, etc.), deploy to Zapier, and an agent watches every run. If a run fails, the agent works out a fix and tests it. Then it's your call whether it auto-deploys small fixes or waits for your OK. Either way, the bigger stuff will ask you first. That means a new action or branch, anything touching authentication, or a fix the agent isn't confident in. In the example below featuring @TweetAnnaMarie, a support ticket came in with no assignee, so the next step crashed. The self-healing agent automatically added a branch that skips the Slack ping when nobody's assigned, so the reply draft still gets logged. Self-healing is part of Next Gen Zaps, which are in early access and free for a limited time on Professional, Team, and Enterprise plans. Waitlist is open: utm.io/us4Cp