Perplexity Adds Grok 4.5 as Orchestrator Model for Computer Pro and Max Users.

Perplexity has added Grok 4.5 as an orchestrator model option for Computer Pro and Max subscribers, announced on LinkedIn this week. The addition gives users another choice for how Computer routes and manages complex queries across multiple AI models.

The company positions Grok 4.5 as both a performance and cost improvement. According to internal testing on WANDR, their evaluation framework, Grok 4.5 scored higher than five other orchestrator configurations while costing roughly half as much as Opus 4.8. That benchmark matters because orchestrator models sit at the center of Computer's architecture, deciding which specialized models handle specific parts of a user's request.

Perplexity has been steadily expanding its orchestrator lineup. The company recently restored Claude Fable 5 after temporarily removing it, and has added Claude Sonnet 5 for Pro and Max subscribers. The pattern suggests Perplexity is testing different orchestrator strategies rather than settling on a single default.

Orchestrators matter more in Computer than in traditional AI chat interfaces. When a user asks Computer to research a topic, write code, or pull data from multiple sources, the orchestrator decides which models handle research, which write the final output, and how to combine results. A cheaper, faster orchestrator can make the entire workflow more efficient without sacrificing quality.

Grok 4.5's cost advantage is notable. At half the price of Opus 4.8, it could make Computer more economical for teams running high-volume workflows. The WANDR scores indicate it's not a budget compromise, the model outperformed existing options in Perplexity's internal testing.

Computer has evolved into a multi-model platform with several layers. Beyond orchestrators, it includes specialized legal integrations, a Brain memory system that retains context across tasks, and Deep Research for complex queries. Adding Grok 4.5 as an orchestrator gives users more control over how these components work together.

The announcement doesn't specify whether Grok 4.5 will become the default orchestrator or remain an optional choice. Given Perplexity's recent pattern of offering multiple orchestrator models, it's likely available as a selectable option rather than a replacement for existing configurations.

For Pro and Max subscribers, the update is live now. Perplexity hasn't shared detailed WANDR benchmark data publicly, so users will need to test Grok 4.5 against other orchestrators in their own workflows to see if the performance and cost claims hold up in practice.