Perplexity’s annualised revenue reportedly rose from below US$250 million in January to more than US$750 million by August, with its Computer agent helping drive the jump. The figures point to a company increasingly selling delegated work and enterprise use, not just AI answers with citations.
Perplexity Watch analysis
What happened
AI Magazine reports that Perplexity’s annualised revenue passed US$750 million in August, citing The Information; it also cites the Financial Times for a US$450 million annual recurring revenue figure in March. These are reported figures, not audited results in the evidence supplied.
The article says Computer, launched in February, orchestrates 19 AI models to research, write and code. It reports that revenue rose 50% in the month after launch. Perplexity says PayPal employees run more than 74,000 queries a week, while Ramp reports a 47% average productivity gain across its go-to-market teams. Those customer figures and results are company-reported, not independent measures of the wider market.
The AI Magazine report also describes a three-year US$750 million Azure agreement and Perplexity’s presence in Samsung’s Galaxy AI on the Galaxy S26. Together, the examples suggest a business spreading across enterprise software, cloud infrastructure and consumer devices.
Why it matters
Perplexity’s pitch began with finding answers and showing sources. The reported growth suggests another ambition: coordinate several models and have them complete multi-step work. That could make model choice less important to customers than the service that routes tasks and connects them to company data.
The distinction matters. A fast-growing revenue estimate is a signal of commercial traction, not proof that every customer is getting the claimed productivity gains or that the growth will last. Still, the reported figures make Perplexity’s enterprise and agent strategy much harder to dismiss as a search-box side quest.
Our read
The interesting shift is from “which model answered?” to “which system got the job done?” If Perplexity can make that orchestration reliable and useful, it has a broader business than a Google challenger. The numbers are striking; the next test is whether they translate into durable customer use rather than a short burst of launch enthusiasm.
What to watch
- Whether Perplexity publishes further revenue figures with clear definitions and comparable periods.
- Whether reported enterprise usage and productivity gains persist across more customers.
- How Computer’s model orchestration works in practice as products and partnerships expand.
Discussion spark: If an AI service gets better by routing work across rival models, should customers care who made the underlying models, or mainly whether the service delivers reliable results?
Sources and evidence
- Perplexity: AI Answer Engine Tops US$750m Run-Rate – AI Magazine (5 November 2026, 15:39 UTC)
Independent WittyWires tracker for public updates about Perplexity. Not affiliated with or endorsed by Perplexity; this is not an official account.