career · career
Cursor PM salary by level
No verified PM salary data exists for Cursor because, until mid-2025, Cursor had essentially no PMs. As of June 2026, Anysphere lists exactly three PM openings against roughly 300 employees. The company crossed $1B ARR by November 2025 and $2B ARR in early 2026 (the fastest SaaS ARR ramp on record) with an engineering-led org that shipped product without a traditional PM function. What follows is the most specific picture available: what the comp structure likely looks like, what the equity actually means at a $29.3B private company, and what differentiates these three roles from any other PM job in the market.
Why there is no public data
Cursor does not disclose salary ranges on its careers page. No state-filing data covers Anysphere yet. Levels.fyi has one data point for a software engineer, not a PM. The engineer compensation that surfaces for “Cursor PM salary” searches ($808K to $1.28M TC with $200K to $300K cash base, the rest in equity) is real for engineers and is a structural ceiling, not a floor, for PM comp.
The three live JDs (Product Manager, Product Manager Agent Harness, Product Manager Cloud Agents) give the clearest signal of what Cursor needs: high-autonomy, technically grounded PMs who can define the behavior boundaries of autonomous coding agents. Not backlog managers. Not roadmap writers.
Estimated salary by level
These ranges are estimates derived from: (1) Cursor engineer cash comp ($200K to $300K base), discounted to reflect PM-vs-SWE market rates at AI-native companies; (2) comparable verified PM comp at Anthropic ($220K to $385K base) and OpenAI (similar structure, higher headline); and (3) general private AI-lab PM benchmarks.
| Role | Estimated base | Estimated TC (paper equity included) |
|---|---|---|
| PM (mid-level) | $180K - $230K | $300K - $600K |
| Senior PM | $220K - $280K | $500K - $1.2M |
| Staff PM | $270K - $340K | $800K - $2M+ |
The TC numbers are wide because they depend almost entirely on equity grant size, grant timing relative to the valuation, and when (if) a liquidity event occurs. On a cash-equivalent basis (what you can actually spend), the honest range for a senior PM joining now is $250K to $400K. That is competitive but below OpenAI ($350K to $500K TC) and roughly in line with Anthropic senior PM ranges ($460K to $625K TC), once illiquidity adjustments are applied to both.
What the equity actually is
Cursor raised $2.3B in November 2025 at a $29.3B valuation, totaling $3.4B raised. The company has $1B in cash reserves and has stated it has no plans to IPO soon as of December 2025.
A PM equity grant at Cursor will typically represent 0.02% to 0.15% of the company depending on level and timing. At a $29.3B valuation, 0.05% is $14.65M in paper value on a 4-year vest schedule. That is a real number. It is also entirely illiquid with no confirmed liquidity event.
The questions that determine whether the equity is worth the discount:
- Does Cursor IPO or enter a structured secondary program before your equity fully vests?
- What does a $29.3B entry valuation mean for upside? The company needs to 3x to 5x from current to produce the kind of returns that make paper equity worth the wait at the original grant size.
- Is there a secondary tender program? Some late-stage privates run annual tender offers to give employees partial liquidity. Cursor has not disclosed one.
If you cannot hold illiquid equity for 3 to 5 years on the cash comp alone, the offer math changes materially. The base is your floor.
How it compares to Anthropic and OpenAI
| Company | Approx. TC (senior PM) | Equity structure | Liquidity path |
|---|---|---|---|
| Cursor | $500K - $1.2M (paper) / $250K - $400K (cash-equivalent) | Private stock options or grants | No IPO timeline confirmed |
| Anthropic | $460K - $625K TC | Double-trigger RSU | IPO expected late 2026 to 2027 |
| OpenAI | $750K - $1.1M TC | RSU / PPU (legacy) | Tender-dependent, structured secondary |
Anthropic and OpenAI have more mature equity markets: Anthropic filed confidentially for an IPO, and OpenAI runs structured tender offers with investor-backed secondary liquidity. Cursor at this stage has neither. The TC headline at Cursor can look larger than Anthropic or OpenAI on paper; on a cash-plus-discounted-equity basis, it is lower.
The role is not a standard PM job
The CPO at Cursor is Sualeh Asif, a co-founder, former IMO mathematics representative, and former IBM neural machine translation engineer. He is not a traditional product hire, and the org reflects that. Cursor ran to $1B ARR because engineering at Cursor can ship directly: feasibility was never the constraint. The PM role that now exists is about identifying what developers will actually pay for and find genuinely lovable in an agentic coding context, where any feature is buildable in days but most features are obnoxious, untrustworthy, or irrelevant.
The three open roles signal specific needs: Agent Harness (how developers stay in control when the agent acts autonomously), Cloud Agents (agentic workflows that run without a local IDE), and a general PM role. These are frontier-product problems about trust, control, and interaction design at the boundary of autonomy. The interviews will probe whether you can think about these tradeoffs, not whether you can write a PRD.
What this means for your negotiation
Cursor will not have a public pay band. That means the negotiation surface is different from Anthropic or Meta, where band data is available and comparable offers are well-documented.
Anchoring points that matter:
- Name a specific competing offer with a cash-equivalent TC. Anthropic and OpenAI offer letters are the most credible comparisons because the roles are structurally adjacent.
- Push on grant size and cliff, not just the headline valuation multiple. At a $29.3B entry valuation, the grant size (as a percentage of the company) matters more than the paper dollar figure.
- Ask explicitly whether Cursor has run or plans secondary tender programs. The answer changes the illiquidity calculation.
- Sign-on bonus is the mechanism for bridging unvested equity you are forfeiting elsewhere. Bring a documented forfeiture number.
For the full negotiation framework, see negotiate equity, not base. For direct comparison, see Anthropic PM salary and OpenAI PM salary.
The 2026 lens
Cursor’s PM hiring in 2026 is a direct consequence of the viable/lovable problem at the frontier. Feasibility is no longer the constraint: Cursor’s own engineers proved that by shipping $2B ARR with almost no PM function. What is left is the harder question: which agentic behaviors do developers actually trust and pay to keep, and which ones are intrusive enough to drive churn? A PM at Cursor is not managing a roadmap. They are defining the interaction contracts for autonomous coding agents at a company that 50,000 businesses depend on. The upside in the equity reflects that bet. So does the illiquidity.
For a broader view of what this shift means for PM comp at frontier labs, see frontier lab comp decoded.