Perplexity is quietly building an AI coding tool to take on Cursor and Claude Code

Perplexity, a San Francisco-based AI search startup, is stepping into the competitive landscape of AI coding tools. The company has developed an internal AI coding assistant, currently codenamed 'Teammate,' which has been in use by its engineers since May. Sources indicate that a public launch could be on the horizon, although the exact timeline remains uncertain. Previously valued at $20 billion during its last funding round, Perplexity originally focused on creating an AI-powered search engine aimed at competing with giants like Google. However, with the potential introduction of Teammate, the company may soon find itself in direct competition with established players in the AI coding realm, such as Cursor, Anthropic, and OpenAI. Teammate is designed to manage software projects from inception to completion. According to an internal announcement, the tool is tailored for long-term engineering tasks, including project ownership, issue investigation, and service monitoring. While Perplexity has not publicly commented on the tool, its engineers have reportedly utilized it for tasks like bug identification in internal systems. Interestingly, Teammate is model-agnostic, meaning it is not reliant on any specific chatbot technology. Perplexity's Chief Technology Officer, Denis Yarats, has been encouraging the engineering team to leverage AI in their coding practices. Just weeks before the internal rollout of Teammate, Yarats communicated that by the end of the year, software engineers should transition away from traditional coding methods to embrace AI solutions. In defense of AI coding tools, Yarats dismissed concerns regarding the quality of AI-generated code, stating that as long as the output meets quality standards, issues of 'slop'—poor-quality code—should not be a concern. This proactive stance highlights Perplexity's commitment to integrating AI advancements within the software development process.