Nine months after raising a $52 million Series B, Graphite got scooped up by Cursor in a deal that reportedly valued the company "way over" its $290 million last valuation. No exact number was disclosed, but the acquisition was financed with a mix of cash and equity, and it came just one month after Cursor announced it had crossed $1 billion in annualized revenue.
If you're an engineer watching the AI dev tools space, this one's worth paying attention to. The acquisition itself wasn't shocking, but the signal it sends about how software gets built from here is hard to ignore.
Graphite built its reputation on a deceptively simple idea: stacked pull requests. If you've worked at Google, you already know the workflow. Instead of waiting for one PR to get approved before starting the next dependent change, you stack them, working on multiple related changes simultaneously while reviews happen in parallel.
It sounds small, but it’s not. For engineering teams shipping fast, the review bottleneck is real. You write code in minutes with Cursor or Copilot, then sit around waiting for a teammate to approve it. Graphite attacked that exact friction point, and it worked. Before the acquisition, the company had tens of thousands of engineers across 500+ companies using the platform, including Shopify, Snowflake, Figma, Perplexity, and plenty of others.
On top of the stacked PR workflow, Graphite layered in AI-powered code review that catches security vulnerabilities, performance issues, and style violations before code ever hits production. They also built Diamond (now rebranded as Graphite Agent), an AI tool that auto-generates fix suggestions and summarizes pull requests so reviewers can move faster.
Cursor has been the breakout story in AI coding. Founded by four MIT grads in 2022, the company went from zero to a $29.3 billion valuation and $1 billion ARR in about three years. Their AI-powered code editor lets developers write and navigate code using natural language, and it's become the default tool for a fast-growing segment of the engineering world.
Cursor's CEO Michael Truell has talked about this openly. Writing code is only half the job. The rest of it (reviewing, testing, merging, deploying) looks pretty much the same as it did three years ago. AI keeps making the writing part faster, which means engineers are spending a growing chunk of their day on everything that comes after.
That's exactly the gap Graphite fills. Cursor owns the writing side and Graphite owns everything that happens after, from reviewing and approving to merging code into production. Together, they cover the full development lifecycle. It's a clean strategic fit, and both companies share investors in Accel and Andreessen Horowitz, which probably made the deal smoother.
For now, Graphite is staying independent, same brand, same product, same team. But throughout 2026, the plan is to build deeper integrations between the two platforms. Smarter, context-aware code reviews that understand how a team actually writes code. Tighter connections between local development and pull requests. And, as Cursor put it, "some more radical ideas we can't share just yet."
If you're already using both Cursor and Graphite, your workflow is about to get better. If you're using one but not the other, expect the combined product to become harder to ignore.
But zoom out a little and the bigger picture matters more. The AI coding market was valued at $1.51 billion in 2024 and is projected to hit nearly $9 billion by 2032. Every major player, including GitHub Copilot, OpenAI, Anthropic, Windsurf, is fighting for developer attention. The competition is fierce and the tools are converging fast.
What makes the Cursor-Graphite combination interesting is that it's not just about generating code. It's about owning the full stack of how code gets from an engineer's brain to production. That's a much stickier product. If your editor, your reviewer, your merge tool, and your deployment pipeline are all talking to each other with shared AI context, switching costs go through the roof.
For individual engineers, this is mostly good news. Better tools, faster workflows, less time waiting on reviews. For engineering leaders, it's a signal to start thinking about how your team's toolchain fits together, because the era of stitching together six different point solutions is probably ending.
Graphite's journey is a fascinating case study in startup timing. The company was founded in 2020, well before the current AI boom. They started with a developer productivity tool (stacked PRs) and layered AI on top as the technology matured. By March 2025, they'd raised $52 million from a murderer's row of investors, including Accel, a16z, Anthropic's Anthology Fund, Menlo Ventures, Figma Ventures, and Shopify Ventures.
Then, nine months later, they got acquired for a premium. That's a recognition that sometimes the best path forward is joining forces with the company that complements you most naturally.
Graphite co-founder and CEO Merrill Lutsky put it well: "We've long dreamed of connecting the surfaces where we create, collaborate on, and validate code changes." Joining Cursor accelerates that timeline by years.
For the broader startup ecosystem, the acquisition reflects a pattern we're seeing more of in AI: consolidation around platforms. The standalone AI tool era is giving way to integrated suites. Companies that do one thing brilliantly are getting absorbed by companies trying to own the full workflow. It's just the natural evolution of a market that's moving incredibly fast.
Cursor acquiring Graphite is just a preview of where the entire software development industry is going. AI is compressing the time it takes to write code, which means the bottleneck is shifting downstream to review, testing, and deployment. The companies that figure out how to AI-power that entire chain, not just the code generation part, are the ones that will define how engineering teams work for the next decade.
Graphite built one of the best code review experiences in the market. Cursor built one of the best code editors. Together, they're making a bet that the future belongs to whoever owns the full loop.
If you're an engineer, this is worth watching closely. And if you're building at a startup that touches developer tools, take notes, because the playbook is being written in real time.