Via cnet.com
The AI search company's latest update lets users query multiple frontier models simultaneously, a capability that could reshape how institutional investors conduct research.
Perplexity AI integrated its Model Council feature directly into Perplexity Computer, its agentic digital worker platform, creating an AI-powered research analyst that can query multiple frontier models at once and synthesize their findings into a single coherent output.
The update went live on March 6, 2026, and includes models GPT-5.4, Claude Opus 4.6, and Gemini 3.1 Pro.
What Model Council actually does #
Model Council runs queries across three frontier models simultaneously, then consolidates their outputs through a synthesizer layer. The synthesizer identifies where the models agree, where they disagree, and what unique insights each one surfaces.
Model Council originally launched as a standalone tool on February 5, 2026. Perplexity Computer followed on February 25, 2026, designed as an agentic digital worker capable of handling complex, extended workflows. The March 6 integration married the two, turning multi-model deliberation into a native capability of the computer agent rather than a separate product users had to switch between.
The update also shipped customizable skills, voice modes, and the ability to choose orchestrator models for different workflows. The full feature set is available exclusively to Max and Enterprise subscribers.
Why this matters for finance and crypto markets #
Perplexity’s official announcements and coverage contain no mention of crypto tokens, protocols, or related market implications. The research context is AI capability rather than blockchain integration.
The competitive landscape #
OpenAI, Anthropic, and Google each want users locked into their respective ecosystems. Perplexity is positioning itself as a layer that sits above individual models, letting users benefit from all of them. OpenAI will never route your query to Claude.
Perplexity Computer can manage extended workflows, meaning it can handle complex, multi-step tasks without human intervention between steps. The addition of voice mode enhances user interaction with the platform.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our