About
Real answers from AI models to one fixed list of prompts, run independently and to be paid for by advertising.
What it is
ModelLineup puts the same 24 prompts to each model and shows every answer in full, next to plain checks, time, tokens and cost. The list never changes, so the answers can be set beside each other. A score is the share of checks an answer passes.
It is free and needs no account. There is no prompt box: you read answers that were collected in advance.
Who it is for
Anyone who uses Claude for real work, such as building a page, fixing code, planning a feature or writing a message, and wants to see how the models differ before relying on one. Each model page shows every answer next to the checks it passed and missed, so you can judge from the answers themselves rather than from a single number.
Who chose the questions
The prompts and their checks were drafted with the help of AI models and fixed before the first run, and they do not change. The prompts page describes each prompt in plain words and lists its checks, and the method page says how the answers were collected and marked.
Who runs it
ModelLineup is operated by Inyonitech Solutions (Pty) Ltd, an independent company registered in South Africa.
It is independent. It is not affiliated with, endorsed by or sponsored by Anthropic. Claude, Fable, Opus, Sonnet and Haiku are trademarks of Anthropic, PBC, named here only to say which model wrote each answer. The site shows no logo of the company that makes the models.
It is to be paid for by advertising, and by nothing else; no advertisement is shown yet. There is no premium tier, no data to sell and no sponsored content. Advertisements will never be placed beside an answer.
Corrections
If a check, a score or a sentence is wrong, write to hello@modellineup.com. Corrections are listed on the what’s new page.