From Copilots to Autonomous Coding Agents
Cognition’s latest funding round highlights a shift in software development, where autonomous coding agents are emerging as systems that can independently plan, write, test, and deploy production code with minimal human oversight, moving far beyond autocomplete-style AI assistants and into the role of full software engineering teammates. Cognition, founded in 2023, built its reputation on Devin, an AI software engineering agent that does more than suggest snippets. Devin can plan tasks, read documentation, run tests, debug failures, and interact with developer tools end to end, so work that once required human engineers can now be automated as a continuous workflow. This is different from earlier AI coding assistants that mostly helped humans type faster. The new model treats the agent as a unit of labor in its own right, capable of owning tickets and projects instead of only supporting them.
A $1B Round at $26B Valuation Changes the Market Narrative
Cognition has raised more than USD 1 billion (approx. RM4.6 billion) at a USD 26 billion (approx. RM119.6 billion) valuation, more than doubling its previous USD 10.2 billion (approx. RM46.9 billion) mark in under a year. The round was co-led by Lux Capital, General Catalyst, and 8VC, with participation from Founders Fund, Ribbit Capital, Atreides Management, and other existing and new backers. According to TechPortal, this brings Cognition’s total capital raised to over USD 2.5 billion (approx. RM11.5 billion). The scale and speed of this AI coding startup funding signal that investors are not treating Devin as a short-lived hype product. Instead, they are pricing Cognition like a platform that could command long-term software and infrastructure budgets, similar to how cloud providers became core enterprise spend rather than side tools for developers.
Devin AI Revenue and Enterprise Proof Points
Devin’s financial traction is one of the clearest reasons behind the valuation surge. Cognition reports an annualized revenue run rate of about USD 492 million (approx. RM2.26 billion), up from roughly USD 37 million (approx. RM170 million) during the same period last year, driven mainly by enterprise AI adoption. Cognition says enterprise usage of Devin has grown more than tenfold since the start of the year, and customers use it as a full engineering teammate rather than a helper. Large organizations including Citi, Mercedes-Benz, Goldman Sachs, Dell Technologies, Santander, the United States Army, and the United States Navy now treat Devin as a core part of their development flows. Mercedes-Benz, for example, cut a legacy modernization project from eight months to eight days, while Itaú Unibanco reports that Devin automatically fixes 70% of security vulnerabilities.
Why Investors Prefer Agents Over Copilots
Investor enthusiasm is not about code suggestions; it is about workflow ownership. Devin shows that autonomous coding agents can plan tasks, call tools, and close tickets with much less human coordination than copilot-style assistants. Cognition’s CEO Scott Wu argues the future of AI coding will depend on systems that combine multiple models and tools, not a single large model, and Devin is built around that idea. This aligns with a wider shift: backers like Lux Capital and General Catalyst are underwriting a world where software engineering becomes a pipeline of machine-run tasks supervised by humans, instead of human-run tasks slightly sped up by autocomplete. With rivals such as OpenAI and Anthropic expanding their own coding agents, this round also looks like a defensive move: investors are locking in a leading independent agent lab before the market consolidates.
The Emerging Playbook for Enterprise AI Adoption
Cognition’s customer stories outline how autonomous agents can be woven into existing organizations without replacing everything at once. Systems integrators such as Infosys and Cognizant have plugged Devin into project delivery workflows, turning it into a force multiplier for existing engineering teams. Fast-growing startups like Exa, Modal, Eight Sleep, and OpenRouter rely on Devin to handle large parts of their software workflows, showing that autonomous coding agents are not only for massive incumbents. Cognition positions itself as an independent agent lab that works with various foundation models, letting enterprises pick the best model for each task while keeping Devin as the orchestration layer. As AI usage scales, organizations are focusing on price-to-performance rather than raw model size, and Devin’s rapid revenue growth suggests that this agent-centric model is starting to win real budget commitments.