LinkedIn has launched a new feature called Crosscheck for premium subscribers, enabling users to test competing AI models from firms such as Anthropic, Google, OpenAI, and Microsoft at no cost.
Crosscheck operates as a “blind taste test” for AI models, where users receive two distinct answers generated by different models in response to a prompt. After selecting their preferred answer, users can see which models provided the responses. LinkedIn’s Chief Product Officer, Hari Srinivasan, noted that Crosscheck is still in the early stages of development and aims to enhance speed and expand the range of models and question types.
The feature includes models from multiple providers, including Anthropic, Google, MoonshotAI, Mistral, and Amazon. Additionally, Crosscheck will feature a leaderboard that monitors user ratings of various models across different sectors.
Crosscheck is limited to text-based prompts and does not allow for image generation, file uploads, or access to advanced tools available on native AI platforms. Users benefit from unlimited text interactions, eliminating worries regarding token limits or additional subscription fees.
LinkedIn will also share anonymized usage data with AI companies to assist them in understanding model performance among various occupational categories. “Anonymized data is shared with model builders to help them understand how their models are performing amongst different occupations,” LinkedIn explained, emphasizing that no personally identifiable information is shared.
Initially, Crosscheck is available only to LinkedIn Premium subscribers in the United States, with plans to extend the feature to additional countries and free users in the near future.
