Factory Reaches $1.5B Value Scaling Autonomous Droids

AI coding startup Factory raised $150 million at a $1.5 billion valuation to scale its autonomous software development platform. The Series C round was led by Khosla Ventures and brings Keith Rabois to the company’s board of directors. For engineering teams comparing AI agent frameworks, Factory’s rapid growth marks a structural shift from basic completion tools toward fully autonomous, end-to-end development agents.

Codebase Context and Multi-Agent Architecture

Factory operates through specialized agents called Droids that execute complex missions across the software development lifecycle. These agents navigate multiple surfaces simultaneously, including IDEs, terminals, CI/CD workflows, and communication platforms like Slack and Teams.

The system relies on two proprietary technologies to manage scale and legacy architecture. HyperCode generates multi-resolution representations of large codebases to map architectural dependencies. ByteRank handles information retrieval, pulling precise context into the agent’s working memory. If you evaluate and test AI agents in production, you know context retrieval is the primary bottleneck for wide-scale refactoring.

Factory bypasses vendor lock-in by maintaining a model-agnostic infrastructure. Enterprises can route tasks through OpenAI’s GPT series, Anthropic’s Claude, or DeepSeek models based on latency requirements and compliance constraints.

Benchmark Performance and Desktop Integration

Factory’s Droids score 19.27% on the SWE-bench Full evaluation and 31.67% on SWE-bench Lite. In September 2025, the platform claimed the top position on Terminal Bench with a score of 58.75%.

The company released a native desktop application for macOS and Windows just days prior to the funding announcement. The client supports multi-agent sessions and native Model Context Protocol integration. This application allows teams to render dynamic AI-native visualizations directly in the workspace, including Mermaid diagrams and flame charts.

Enterprise Deployment Constraints

Factory specifically targets highly regulated sectors, including systemically important banks, healthcare providers, and national security organizations. The platform supports cloud-managed, hybrid, and fully airgapped deployments. This infrastructure allows large engineering divisions to reduce LLM API costs while maintaining strict data governance over proprietary code.

Publicly named customers include Morgan Stanley, Nvidia, Adobe, and Palo Alto Networks. Teams at these organizations use Droids for heavy-lift tasks ranging from incident response to automated COBOL-to-modern-language migrations. Revenue has doubled every month over the past six months.

Engineering leaders should review their current automated testing and CI/CD pipelines for agent compatibility. Moving to agent-native development requires deterministic testing environments and well-documented repository structures. Prepare your infrastructure to support agents that commit code and manage pull requests autonomously.