Factory has just announced the completion of a $150 million Series B round, valuing the company at $1.5 billion. The round was led by Khosla Ventures, with participation from Sequoia, Insight Partners, and Blackstone.
Founder Matan Grinberg is a PhD candidate in physics at UC Berkeley. In 2023, he sent a cold email to Sequoia partner Shaun Maguire; the two talked physics and investment, and Factory was born.
Three years later, Morgan Stanley, EY, and Palo Alto Networks are paying customers.
What Factory does
It is not Cursor, not Claude Code, not Copilot.
Factory builds enterprise-grade AI coding agents — automated workflows designed for large engineering teams, not code completion for individual developers.
The core scenario: a large company with hundreds of engineers, a massive codebase, complex context, compliance requirements, and internal tools. Tools like Cursor are optimized for individual developers; Factory approaches from the enterprise side, solving the problem of deploying AI coding at scale within an organization.
Why enterprise customers buy this
What kind of organizations are Morgan Stanley and EY? Finance and consulting. These two industries share common traits:
- Code changes require approval processes
- Extremely high compliance and security requirements
- Large engineering teams but complex internal systems
Letting engineers use Cursor directly? Possible, but enterprises worry: which model is the code running on, is data leaking, who reviewed it, where is the process?
Factory's positioning is to add a management layer to these concerns — access control, audit logs, integration with internal enterprise tools. This is not just a feature; it is a governance framework.
Differences from Cursor and Claude Code
Factory's founder says their key advantage is the ability to switch between different foundation models — for example, between Anthropic's Claude and DeepSeek.
But Cursor can do this too.
The real differentiation may lie in the starting point: Cursor evolved from a code editor; Factory was built from the start as an enterprise task orchestration platform. Different starting points lead to different use cases.
| Product | Primary users | Core scenario |
|---|---|---|
| Cursor | Individual developers / small teams | Code writing and refactoring |
| Claude Code | Individuals / small to medium teams | Terminal tasks, agent execution |
| Factory | Enterprise engineering departments | Multi-person collaboration, compliance workflows, large repositories |
Competition is not about one eliminating the other, but about who can first secure a position in the enterprise budget.
Blackstone's participation is worth a second thought
Khosla leading Factory is standard tech VC behavior, but Blackstone's involvement is interesting.
Blackstone is not a pure tech VC; it is a major PE firm. Its entry into the AI coding space represents a judgment on the restructuring of the enterprise software market, not just a bet on a technology trend.
PE and VC see different things: PE looks at future cash flow in this market and whether traditional enterprise software procurement processes will be replaced by AI. Blackstone said "yes" by joining this round.
This follows the same logic as Blackstone's earlier investment in Anthropic — first bet on the model layer, then on the application layer.
The real battlefield for AI coding in 2026 is the enterprise
The story of AI coding tools was told in 2024-2025, with developer tools as the protagonist. In 2026, the second act is unfolding, and the protagonist has become the enterprise.
What's the difference? Developers pay individual subscriptions, at most a few dozen dollars a month. Enterprises pay annual contracts, often starting at six figures. Factory is getting the latter.
The fact that Morgan Stanley and EY are already paying is more convincing than any benchmark.
Source: CocoLoop, Factory hits $1.5B valuation to build AI coding for enterprises (TechCrunch)