COGNITION HITS $445M RUN RATE IN 18 MONTHS
AI DESK■ 2 MIN READ
TUE, MAY 12, 2026■ AI-SUMMARIZED FROM 1 SOURCE ▸ TIMELINE
Cognition, the AI startup behind the Devin coding agent, has reached a $445M annual revenue run rate in its first 18 months of operation. CEO Scott Wu, whose background includes competitive mathematics, discusses the company's rapid growth and AI development strategy.
Cognition has achieved remarkable commercial traction with its Devin AI coding agent, reaching $445M in annualized revenue within 18 months of launch. The milestone reflects strong enterprise adoption of AI-powered development tools.
Scott Wu, Cognition's founder and CEO, brings unconventional credentials to the role. His background in competitive mathematics shaped his approach to problem-solving and product development. This foundation influences how Cognition approaches AI architecture and algorithm design.
Devin, Cognition's flagship product, represents a shift toward autonomous AI agents capable of handling software engineering tasks. The agent can read, write, and execute code, debug applications, and manage development workflows. Early adopters have used Devin for tasks ranging from feature development to infrastructure management.
The company's revenue trajectory outpaces typical enterprise software adoption curves. Cognition attributes this to immediate utility—customers deploy Devin to augment existing engineering teams rather than replace them. This positioning has resonated with enterprises looking to accelerate development cycles.
Wu's mathematical background connects to Cognition's technical approach. Competitive mathematics emphasizes elegant solutions, constraint optimization, and rigorous proof—principles applicable to AI training and deployment. Wu spent years in competitive circuits before transitioning to technology entrepreneurship.
Cognition operates in a crowded AI agent space, competing with larger players and well-funded startups. The company's early focus on developer tools provided clear ROI metrics and faster sales cycles compared to broader enterprise AI applications.
The $445M run rate valuation hasn't been publicly disclosed, but the revenue milestone suggests significant enterprise penetration. Large software companies and tech-forward organizations represent primary customers.
Looking forward, Cognition faces questions about maintaining growth rates as the AI agent market matures. The company continues iterating on Devin's capabilities, particularly around complex multi-step engineering projects and cross-team coordination.
■ SOURCES
► Techmeme■ SUMMARY WRITTEN BY AI FROM THE LINKS ABOVE
■ MORE FROM THE AI DESK
Thinking Machines Lab announced a research preview of interaction models that process real-time dialogue natively, enabling continuous back-and-forth collaboration between users and AI without external scaffolding.
3H AGO— AI Desk
Thinking Machines, founded by former OpenAI CTO Mira Murati, announced it is developing "interaction models" designed to enable natural human-AI collaboration through simultaneous audio, video, and text processing.
3H AGO— AI Desk
Agentic inference differs fundamentally from today's language model inference, shifting computational priorities away from speed since human latency constraints no longer apply.
9H AGO— Industry Desk
Baidu's new Ernie 5.1 model achieves competitive performance with just six percent of typical pre-training costs by using a third fewer parameters than its predecessor. The model ranks fourth globally on Search Arena leaderboards, behind Claude Opus variants and GPT-5.5 Search.
9H AGO— AI Desk