LinkedIn has launched a new feature called Crosscheck that allows premium subscribers to test AI models from various companies including OpenAI, Anthropic, Google, and Microsoft without concerns over token limits or additional subscriptions.
Crosscheck enables users to engage in what LinkedIn describes as a “blind taste test,” where users receive two AI-generated responses based on a prompt. Users can only identify which AI model produced each answer after making their selection, facilitating an unbiased evaluation of the models.
Hari Srinivasan, LinkedIn’s Chief Product Officer, stated that Crosscheck is an “early product” with ongoing developments planned to enhance its speed and broaden model support. Initial tests have shown compatibility with multiple models from companies including MoonshotAI and Mistral.
The feature includes a leaderboard to track user ratings for each AI model across different industries, fostering a competitive atmosphere. However, Crosscheck currently supports only text-based prompts and does not allow for image generation or file uploads.
Subscribers are permitted to conduct unlimited text-based chats without facing token limits or hefty subscription fees. LinkedIn will share anonymized data with the AI companies, helping them to analyze model performance among different occupational categories while ensuring no personally identifiable information is disclosed.
Currently, Crosscheck is available exclusively to LinkedIn Premium subscribers in the United States. LinkedIn has plans to expand the feature to additional countries and to free users in the near future.
