The timing, as they say in Silicon Valley, was no accident.
When GitHub flipped the switch on usage-based billing for Copilot Business this past June—$19 per user monthly—Cursor unveiled its own pricing shake-up just days earlier, on June 1st. Teams would run $32 per user monthly on an annual plan, or $40 if you paid month-to-month. A new Premium tier landed at $96 monthly when billed annually. Not cheaper, certainly. But Cursor's pitch wasn't about undercutting GitHub on price. It was about selling a fundamentally different vision of how AI belongs in software development.
For the better part of this year, Cursor has been methodically assembling the infrastructure of an enterprise-grade platform—Organizations, security tooling, team collaboration features—the kind of unglamorous plumbing that signals a company no longer content to serve individual developers tinkering on side projects. What's emerged is less an incremental improvement than a direct challenge to Copilot's embedded-assistant model, one that centers on what Cursor calls "agent-first workflows."
Whether enterprises will buy that pitch remains an open question. But the company is forcing the conversation.
Building the Enterprise Stack
The Organizations feature arrived June 3rd, bringing centralized management for companies juggling multiple development teams. Administrators got a single dashboard to wrangle separate budgets, security policies, model access controls, and development sandboxes—table stakes for any tool hoping to land deals with IT procurement departments. According to Cursor's announcement, several customers had already been testing the capability in beta, though the company didn't name names.
It capped a sprint of enterprise-focused releases that had been building since spring. Security Review entered beta in late April, deploying what Cursor describes as always-on security agents—a Security Reviewer and Vulnerability Scanner—that crawl through pull requests and codebases according to policies teams configure themselves. The feature stacks alongside Cursor's SOC 2 Type II certification, the compliance checkbox documented on its security page as of late April.
The subtext of these moves is hard to miss: Cursor wants to sit in the same procurement conversations as GitHub, and it's checking off the boxes required to get there.
The Marketplace Bet
Where Cursor diverges—and where its bet gets interesting—is in its emphasis on shared, reusable AI workflows. The Teams product includes something the company calls a private marketplace, where developers can publish internal plugins, skills, rules, and hooks that become, in effect, shared infrastructure across an organization. It's a different mental model than Copilot's IDE-embedded suggestions or Anthropic's CLI-centric Claude Code, as an InfoQ analysis noted back in mid-April.
This marketplace builds on Cursor's Automations system—"always-on, event-triggered agents," as TechCrunch described them in early March—that can respond to pull requests, Slack messages, or scheduled triggers. Bugbot, Cursor's code review agent that graduated from beta last summer, operates as one such automation. The idea, roughly, is that instead of every developer invoking AI assistance individually, teams build libraries of agents that do recurring work automatically.
Shared Canvases, which launched in late May, extends this collaborative foundation by letting teams share live snapshots of agent-generated work through a dashboard interface. It's available across Pro, Teams, and Enterprise tiers, a deliberate attempt to make collaboration a selling point rather than an enterprise-only feature gated behind the highest price tag.
Expansion Beyond the Desktop
In late June, Cursor entered public beta with an iOS app—a practical, if unsurprising, extension of its "everywhere your team works" positioning. Developers can launch, monitor, and control cloud agents from their phones, complementing desktop, web, and Slack surfaces. It's hardly revolutionary, but it signals ambition beyond the code editor as a destination, toward orchestration across whatever surface happens to be closest.
The company is pushing trial adoption hard, perhaps more aggressively than its relatively modest public profile would suggest. A dedicated landing page offers up to $3,000 in Teams trial credits covering both seats and agent usage—access to Cloud Agents, Bugbot, Automations, Security Review, the works. The credits expire after four weeks, a ticking clock designed to convert curiosity into contracts.
Courting the Enterprise
Cursor has paired its product roadmap with strategic partnerships aimed squarely at large-scale adoption. EPAM announced a collaboration in early January focused on "building and scaling AI-native teams," complete with reference implementations, training programs, and maturity models—the kind of professional services wrapper that helps enterprise software cross the chasm from early adopters to mainstream IT budgets. AHEAD followed in late May with a partnership targeting what it called "enterprise-scale AI-native SDLC transformation," language that reads like it was workshopped in a conference room but signals the same intent.
A security partnership with Chainguard, covered by Axios in late April, aims to steer AI-generated code toward vetted components—a governance feature clearly designed for risk-averse organizations worried about supply chain security and compliance.
The enterprise feature set now reads like a checklist: SSO (SAML/OIDC), SCIM provisioning, role-based access control, privacy mode enforcement, audit logs, repository-level permissions. Usage analytics and admin dashboards give IT leaders visibility into spending and team behavior. It's the kind of infrastructure that doesn't make for exciting product launches but that procurement departments demand.
The Calculation
Cursor's June pricing overhaul introduced separate usage pools for its Composer tool versus third-party API calls, with at least $20 per month in agent usage baked into each seat. The Premium tier offers five times the usage at three times the price, a bet on high-volume power users who'll justify the economics through sheer volume.
Whether enterprises will pay a premium for agent-first infrastructure over GitHub's embedded assistance—familiar, lower-priced, already woven into Microsoft's enterprise fabric—is the bet Cursor is making. The company's wager is that organizations willing to rethink development workflows around autonomous agents will value shared marketplaces, always-on automations, and multi-surface orchestration enough to justify the cost difference.
The aggressive trial program suggests Cursor understands it needs to prove that thesis in practice, not just in pitch decks. Four weeks to demonstrate value, $3,000 in credits to remove friction, and a clock ticking down to conversion. It's a tried-and-true enterprise software playbook. Whether it works against an incumbent as entrenched as GitHub Copilot is another matter entirely.
