---
title: "Greptile Raises $50M at $750M to Check AI-Written Code"
description: "Menlo led Greptile’s $50M Series B. Its credit pricing and Mixpanel’s workflow show what a separate AI code reviewer must prove."
date: 2026-10-25T13:14:36.000Z
source: https://venturecapitaltracker.com/greptile-50m-series-b-750m-code-review
---

# Greptile Raises $50M at $750M to Check AI-Written Code

> Menlo led Greptile’s $50M Series B. Its credit pricing and Mixpanel’s workflow show what a separate AI code reviewer must prove.

Greptile [announced a $50M Series B at a $750M valuation](https://www.greptile.com/blog/series-b) on October 19, betting that checking AI-written software can become a business in its own right. The company says it reviews and tests more than 1% of pull requests opened on GitHub daily. That is a usage claim, not a revenue figure or evidence that those changes are bug-free.

[Menlo Ventures](/fund/menlo-ventures) led the round; Jump Trading, Wing, Operator Collective, Benchmark, Y Combinator, Browder Capital and Initialized participated. Menlo’s Deedy Das is joining the board. The announcement does not specify whether the valuation is pre- or post-money, disclose revenue, or break out any secondary share sales.

The financing follows Greptile’s [Benchmark-led $25M Series A in September 2025](https://www.greptile.com/blog/series-a). The new round is twice that size, but neither amount tells investors how efficiently the software converts usage into profit.

## What customers are paying to check

Greptile reviews proposed code changes before they join a shared codebase. A pull request is the submission developers use to ask teammates to check and merge those changes. Its commercial opportunity is the work between generating software and deciding it is safe to ship.

In a [Greptile-published Mixpanel case study](https://www.greptile.com/customers/mixpanel), platform engineering manager Austin Pray describes testing outside providers alongside an internal reviewer. The home-built version was costly and mixed trivial comments with serious problems. He says Greptile identified a change that would affect a downstream service despite the submission’s claim that it was isolated.

The practical change was earlier machine feedback, leaving human reviewers more time for the purpose and design of a change. This is a vendor-hosted customer account, not an independent controlled test or proof that all dangerous changes are detected.

Greptile’s [pricing page, checked October 25](https://www.greptile.com/pricing), lists Pro at $1 per credit with no seat fees. Base reviews consume one credit, Plus three and Apex ten; credits are shared across an organization. Enterprise contracts are custom-priced.

Illustratively, 100 Base reviews cost $100 at the listed rate, while 100 Apex reviews cost $1,000 before negotiated terms. These are calculations, not customer bills. Review volume and review depth, rather than engineering headcount alone, drive spending. For Greptile, the corresponding unanswered question is how much computing each useful review costs.

## A funded market with difficult substitutes

[CodeRabbit’s $143M Series C](/2026-coderabbit-143m-series-c-agentic-change-management), announced August 12 at a $1.5B valuation, finances expansion beyond review into broader management of software changes. [Qodo’s $70M Series B](/2026-qodo-70m-code-verification), announced March 30, backs enterprise review and governance. Their [company](https://www.coderabbit.ai/newsroom/coderabbit-series-c-agentic-change-management) [announcements](https://www.qodo.ai/blog/qodo-70m-series-b-shift-to-artificial-wisdom/) establish comparable investments, not equivalent products or operating performance.

Menlo’s [portfolio page](https://menlovc.com/portfolio/greptile/) confirms its Series B involvement and describes Greptile’s security-review capabilities; it does not publish a detailed investment argument. VCT’s reading is that buyers may pay a separate specialist when reviewing more machine-generated work consumes scarce engineering attention.

That position still has to beat an internal reviewer and review features supplied by coding-agent vendors. A second AI system can reproduce the first system’s mistakes, and noisy warnings can consume the time they promise to save. The decisive measures are serious defects caught, false alarms, engineer time spent resolving findings and retained spending after the trial. At $750M, Greptile must show that customers keep paying for better judgment as the cost of generating code falls.

**By:** [Venture Capital Tracker](https://venturecapitaltracker.com/editorial-policy)

**Editorial note:** AI tools assisted with research, structure, or drafting. Venture Capital Tracker retains human editorial responsibility for factual accuracy, relevance, and source quality before publication.