Perplexity unveils Model Council to compare answers across AI models: How it works

Perplexity has launched Model Council, a new AI feature that runs queries across three models at once, compares their responses and highlights key insights, helping users verify information, reduce inconsistencies and make better-informed decisions faster.

Govind Choudhary
Updated9 Feb 2026, 09:12 PM IST
Perplexity has introduced a new artificial intelligence feature aimed at improving the accuracy and depth of responses on its platform, allowing users to cross-check answers from several models at once.
Perplexity has introduced a new artificial intelligence feature aimed at improving the accuracy and depth of responses on its platform, allowing users to cross-check answers from several models at once.

Perplexity has introduced a new artificial intelligence feature aimed at improving the accuracy and depth of responses on its platform, allowing users to cross-check answers from several models at once.

Called Model Council, the tool enables users to submit a single query to three separate AI models simultaneously. The system then analyses the outputs and presents a consolidated view, highlighting key insights, areas of agreement and points of difference.

Tackling inconsistent AI responses

As AI systems become more sophisticated, their methods of reasoning and presenting information can vary widely. The same complex question posed to different chatbots often produces noticeably different answers, which can leave users unsure which version to trust.

Perplexity says this inconsistency can be frustrating, particularly for people conducting research or making important decisions. Reading and comparing multiple responses manually is time-consuming and inefficient.

Model Council is designed to address that issue by automating the comparison process.

Also Read | ‘The spark remains human’: Perplexity CEO on why AI lacks curiosity

How it works

Users select three AI models within the platform and submit their question. Each model generates an independent response. Perplexity’s built-in synthesiser then reviews the outputs and compiles them into a structured table.

This summary format shows where the models agree, where their views diverge and flags any unique insights offered by a specific system. Users can also open and examine each model’s full answer separately if they wish to dig deeper.

The company says this approach allows people to grasp the headline information quickly while still retaining the option to verify details.

Also Read | Perplexity AI adds a ‘conversational’ shopping assistant

Focus on accuracy and perspective

Perplexity suggests the feature will be most useful in scenarios where precision and breadth of perspective matter. Suggested use cases include investment research, complex decision-making, creative ideation and fact-checking.

By drawing on multiple AI viewpoints at once, the company hopes users will gain a more rounded understanding rather than relying on a single source.

Availability

Model Council is currently limited to Perplexity’s Max subscribers. The firm has indicated that the feature could expand to its Pro tier in the near future.

About the Author

Govind Choudhary is a Senior Content Producer for Mint with over 04 years of experience covering technology and automobiles. He holds a Master's diplo...Read More

Get Latest real-time updates

Catch all the Technology News and Updates on Live Mint. Download The Mint News App to get Daily Market Updates & Live Business News.

Business NewsTechnologyNewsPerplexity unveils Model Council to compare answers across AI models: How it works
More