Perplexity Launches “Computer” Platform, Bundling Rival AI Models into One $200/Month

Perplexity’s “Computer” platform represents a concrete step toward unified, multi‑model orchestration, a capability that has long been limited to bespoke research frameworks. According to the announcement on Perplexity’s LinkedIn page, the service runs Claude Opus 4.6 as its central reasoning engine while dynamically delegating sub‑tasks to a roster of specialist models—including Gemini for deep web research, Grok for lightweight compute, ChatGPT 5.2 for conversational glue, Nano Banana for image generation, and Veo 3.1 for video synthesis. Each delegated task executes inside an isolated sandbox that provides a full browser instance, a persistent file system, and a suite of tool connectors, allowing the system to persist state across days or weeks without user supervision (The‑Decoder, Feb 25 2026).

The architecture mirrors the “orchestrator” pattern popularized in recent academic work on autonomous agents, but Perplexity has packaged it as a consumer‑grade SaaS offering. Users describe a high‑level outcome—such as “build a full‑stack web app with a PostgreSQL backend and deploy it to AWS”—and the platform automatically spawns sub‑agents to perform research, write code, run tests, and issue API calls. Because each sub‑agent runs in its own secure environment, the risk of “context collapse” or infinite loops—problems that have plagued earlier multi‑agent attempts—is mitigated (Kowshik Jallipalli, Feb 25 2026). The service also supports asynchronous execution, meaning that long‑running jobs can continue in the background while the user is free to issue new prompts or inspect intermediate artifacts.

Pricing is positioned to target professional developers and small enterprises rather than hobbyists. Perplexity bundles the Computer into its “Max” subscription at $200 per month, a price point that reflects both the compute overhead of maintaining multiple concurrent model instances and the value of a single, unified API surface. The company’s rationale, as outlined in the product brief, is that AI models are increasingly specialized; a single model cannot efficiently handle the full spectrum of tasks required for end‑to‑end workflows (The‑Decoder, Feb 25 2026). By aggregating models from competing providers, Perplexity sidesteps the need for users to negotiate separate contracts or manage disparate credentialing, effectively turning the fragmented model ecosystem into a cohesive service layer.

From a competitive standpoint, Perplexity’s move challenges both pure‑play chatbot firms and emerging “agentic” platforms such as Anthropic’s Claude Cowork. While Claude Cowork offers a browser‑based interface, it remains tied to a single underlying model and lacks the sandboxed multi‑tool environment that Perplexity advertises. Conversely, open‑source orchestration stacks like LangChain require users to assemble their own toolkits and handle security concerns themselves. Perplexity’s “Computer” thus occupies a middle ground: a proprietary, turnkey solution that still leverages third‑party models, potentially lowering the barrier to entry for companies that need autonomous workflows but lack deep AI engineering resources.

The launch also dovetails with Perplexity’s broader push into developer‑facing products, most recently a “massive search API” that exposes its web index to external applications (VentureBeat, 2024). By coupling that API with the Computer’s multi‑model workflow engine, Perplexity could enable end‑to‑end pipelines that start with raw web data, transform it through specialized models, and output finished artifacts—all without leaving the Perplexity ecosystem. If the platform scales as advertised, it may force other AI vendors to reconsider the siloed model approach that has dominated the market since 2023, accelerating a shift toward integrated, agent‑centric services.