The journey so far and the road ahead.
Six months ago, we celebrated 1M users on Resend.
Today, we crossed 3,000,000.
Here's a quick recap and what's next.
In just six months, we went from:
This is how we're seeing the path ahead.
Resend is profitable. It has been since last year.
We didn't get there by cutting costs. We never reduced our cloud spend, never trimmed our Datadog bill, and never slowed hiring to hit the number sooner.
We got there by staying focused on the product, the infrastructure, and keeping customers happy. Profitability is the byproduct, not the end goal.
The usual startup advice is to burn cash and grow faster. Those aren't opposites. You can grow fast, move aggressively, and still be profitable.
Profitability isn't about saying no to VCs. The right investors make you better, and ours did. Profitability gives you options: raise more or less, now or later, or not at all. It means you can optimize for the best decision for the product.
We're profitable, we'll remain profitable, and we'll continue to move insanely fast.
The way we approach support is still one of the things that sets Resend apart.
Most hypergrowth companies put AI in front of support to keep ticket counts down. The result? AI becomes a wall between the customer and the team. A wall that you can't reach over.
We're doing the opposite. We're putting the product behind the agents.
try {product()codeExamples()documentation()knowledgeBase()aiAgent()}catch {support()}
We're equipping our AI agents with the same tools that power the app, so the easy questions are answered right away.
And when the answer isn't there, you're one message away from a human, a human who is an engineer, who actually knows about email.
The bar for support goes up, not down. The more we automate the stack for LLMs, the more valuable real humans become since they can answer the difficult questions.
We spent years obsessing over developer experience.
Now we obsess over agent experience, too.
Agents don't need dashboards. They need APIs. They need clean docs, and a way to act without a human in the loop.
Every page on resend.com now renders as Markdown, not just the docs. Our MCP server went from ~8% to ~90% API coverage. We shipped a CLI with 53 commands, so anything you can do in the dashboard, an agent can do in a terminal. And agent skills can guide your entire agent experience, from using the CLI, to the Resend API, to React Email.
Next, we're shipping a remote MCP server and OAuth for agents, so an agent can authenticate and send on a user's behalf directly from Claude, Codex, or Cursor.

Official Resend MCP Server
Our official MCP server with Resend platform coverage, and an official HTTP transport.
resend.com/changelog/mcp
Anyone can prompt their way to a working API, a generic landing page, or a basic dashboard.
When everyone can ship "okay," "okay" stops being worth anything.
From day one, we wanted to ship things that don't only work, but also inspire. True quality is worth copying.
Because we believe there are three types of quality:
Quality Type 3 is the only one that matters to us.
We pursue deep quality not because craft is nice, but because a product that's a joy to use is the one users remember, recommend, and refuse to leave.

Savour
A newsletter by the design team at Resend on taste, craft, and the details that make great products.
resend.com/savour
Who likes stitching together three services to do the same thing? No one.
There should be no need to use multiple email services: just Resend.
We recently shipped Automations, and the early usage is some of the strongest we've seen for a new product.
Now, you can send, receive, broadcast, and automate, all in one place.
All-in-one matters even more in the agent era. For an agent, every extra service is another billing account to manage, another credential to configure, another thing that can break.
Our goal is to continue closing the gaps, so you don't have to.

Automations
Build lifecycle sequences triggered by events in your app.
resend.com/blog/introducing-automations
The last era of software was built for humans clicking buttons. The next one is for developers building with agents.
We're building for both: the developer and the agent in the loop.
We intend to be the email layer underneath all of it.
But we can't do it alone, so come build with us.