Perplexity plans 2028 IPO as CEO unveils vision for multi-agent AI computing

The AI search startup says its public listing plans remain on track regardless of rival IPO outcomes, while outlining an ambitious roadmap for a platform that coordinates multiple AI models to automate complex digital workflows.

Artificial intelligence startup Perplexity is maintaining its long-term plan to pursue an initial public offering in 2028, even as several of the industry's biggest players prepare to test investor appetite for AI stocks through upcoming public market debuts.

Perplexity Chief Executive Officer Aravind Srinivas said the company’s IPO timeline remains unchanged and is not dependent on how other high-profile AI firms perform after listing. His comments come amid growing anticipation around potential offerings from leading AI companies, including OpenAI and Anthropic, as well as other technology firms seeking to capitalize on strong investor interest in artificial intelligence.

Perplexity executives believe that maintaining a disciplined timeline has allowed the company to focus on sustainable growth rather than short-term market sentiment. The company has consistently indicated that a public offering would not be considered before 2028, a strategy executives say has helped build a stronger business foundation.

Srinivas acknowledged that the performance of major AI-related IPOs could influence broader market sentiment toward the sector. However, he emphasized that Perplexity’s long-term plans remain independent of those developments, even while recognizing that successful listings would benefit the wider AI ecosystem.

Betting on a new AI operating system

Alongside its IPO ambitions, Perplexity is positioning itself around a broader vision for the future of AI-powered computing. During an appearance at COMPUTEX 2026, Srinivas outlined the company’s concept of “Perplexity Computer,” a platform designed to coordinate multiple AI models and digital tools within a single workflow.

The system aims to move beyond conventional chatbots by functioning as an intelligent digital worker capable of managing complex assignments. Rather than relying on one model, the platform can distribute tasks across different AI systems depending on their strengths, while balancing factors such as cost, performance, privacy, and efficiency.

Introduced earlier this year, the platform is designed to orchestrate multiple AI agents that can collaborate on projects involving research, content creation, image generation, video production, and other tasks requiring specialized capabilities.

Multi-model approach to AI automation

Perplexity’s platform is built around a model-agnostic architecture that allows it to select the most suitable AI model for specific subtasks. The system can assign research-intensive work to one model, creative generation tasks to another, and reasoning functions to a separate engine, creating an integrated workflow managed by AI agents.

According to the company, users simply describe the outcome they want to achieve, after which the platform breaks the assignment into smaller tasks and executes them through coordinated AI agents.

The strategy reflects a growing trend within the AI industry toward agentic systems that can perform extended, multi-step workflows across different applications. As competition intensifies among AI providers, Perplexity is betting that future computing experiences will be defined not by a single model, but by intelligent systems capable of seamlessly orchestrating many of them.