Perplexity Says xAI's Grok 4.5 Tops Its WANDR Benchmark as Orchestrator in Perplexity Computer

Perplexity has named xAI's newly released Grok 4.5 the top performer on its internal WANDR benchmark when used as the orchestrator model inside Perplexity Computer, positioning it in the same price range as a recently announced custom GLM-5.2 configuration. The claim, tied to Perplexity's agentic research system, arrives days after xAI formally shipped Grok 4.5 and sharpens an intensifying competition over cost-efficient agent orchestration.