en.Wedoany.com Reported - Perplexity is quietly building an AI coding tool to compete with Cursor, Anthropic, and OpenAI—all three of which already have widely used coding tools. The San Francisco-based startup, known for its AI search engine, has developed an internal coding tool codenamed "Teammate," which its engineers have been using since May and may be publicly released later.

According to an internal announcement seen by Business Insider, Teammate is designed to run software projects end-to-end, built for long-term engineering work, handling projects, investigating issues, and monitoring services. Screenshots show that engineers can assign actual tasks to it, such as finding bugs in internal systems. The tool remains model-agnostic and does not rely on any single chatbot, a stark contrast to competitors like Claude Code, which depend on their own models. A Perplexity spokesperson declined to comment, and it is unclear whether or when the product will be released. The startup was valued at $20 billion in a funding round last year, giving it room to experiment.

Weeks before Teammate went live internally, Perplexity CTO Denis Yarats told engineers that by the end of the year or sooner, they should "stop looking at code" and use AI directly. This echoes the shift toward coding assistants within OpenAI. Yarats also pushed back against criticism that AI writes "low-quality code," arguing that as long as the generated code passes quality checks, low-quality code is not a problem.

AI coding is one of the few areas where the technology has truly generated revenue, attracting a wave of well-funded entrants. A model-agnostic tool from a $20 billion search company will add a new perspective to this fiercely competitive battlefield. It should be noted that this report is based on a single source, and Perplexity has not confirmed a release. But the direction is clear: even an AI search company now sees writing code—not just answering questions—as a domain worth capturing.