Generate original ACT Reading passages across all four passage types with a literal-versus-inferential answer key
ACT Reading dropped from 40 questions to 36, but the section actually gained five minutes, up from 35 to 40. Fewer questions and more time means more room per passage than the format used to allow, a real shift in pacing, not just a smaller test. You are an ACT Reading tutor building original passages that match the section's current four-passage structure. Write one passage per type from [PASSAGE_TYPE:select:Literary Narrative,Social Science,Humanities,Natural Science,Mixed across all four], 650 to 750 words each, following the real subject conventions: Literary Narrative asks a reader to infer character and emotion, Social Science tracks a factual argument, Humanities swings between narrative and analytical writing, Natural Science presents dense technical content that rewards careful literal reading over inference. Generate [PASSAGE_COUNT:number:1-4] passages, nine questions after each, testing both what's directly stated and what's implied, matching the real question distribution. Give four answer choices per question. In the answer key, tag each question as either literal or inferential, and note that on the real test, some questions in a set are unscored field-test items mixed in without being labeled, which is exactly why guessing which ones don't count is never a useful strategy, treat every question as if it's scored. Close by naming, based on the mix of missed questions, whether the pattern points to a specific passage type being weaker, or to inferential questions specifically being harder than literal ones regardless of passage.
Range: 1 - 4
Use this prompt anywhere
10,000+ expert prompts for ChatGPT, Claude, Gemini, and wherever you use AI.
Get Early AccessACT Reading dropped from 40 questions to 36, but the section actually gained five minutes, up from 35 to 40. Fewer questions and more time means more room per passage than the older format allowed, a real shift in pacing, not just a smaller test.
This tool builds original passages matching the section's current four-passage structure. Set [PASSAGE_TYPE] to Literary Narrative, Social Science, Humanities, Natural Science, or a mixed set, and choose [PASSAGE_COUNT] for how many to generate, each 650 to 750 words with nine questions after it, testing both what's directly stated and what's implied.
Every answer key tags each question as literal or inferential, since some passage types, Natural Science especially, reward careful literal reading over inference more than the others do. A handful of questions on the real test are unscored field-test items mixed in without a label, which is exactly why treating every question as scored is the only useful approach. Pair this with the ACT English practice generator for the rest of the verbal side, or generate a passage set now in the Dock Editor.
Open the ACT Reading prompt in the Dock Editor, Claude, ChatGPT, or Gemini, then set [PASSAGE_TYPE] to Literary Narrative, Social Science, Humanities, Natural Science, or a mixed set covering all four.
Choose [PASSAGE_COUNT] from 1 to 4 depending on session length.
Watch for both directly stated facts and implied meaning, since the answer key tags each question by which type it is.
A small number of real ACT Reading questions are unlabeled, unscored field-test items, so there's no reliable way to skip effort on any specific question.
Look at whether errors cluster around one passage type or around inferential questions specifically, regardless of passage.
Set [PASSAGE_TYPE] to Natural Science or another specific type to target the passage style causing the most missed questions.
Practice with the current 36-question, 40-minute pace instead of the older 40-question, 35-minute format.
Assign focused practice on one passage type at a time instead of a mixed four-passage set every session.
Use the literal-versus-inferential tags to confirm whether inference specifically, not reading speed, is the real gap.
Discover more prompts that could help with your workflow.
Generate LSAT Reading Comprehension passage sets with an answer key that quotes the exact supporting line for every answer.
Generate TOEFL Writing practice across Build a Sentence, Write an Email, and Academic Discussion tasks with model responses and score-based feedback
Generate GED Math practice questions split into no-calculator and calculator-allowed sections with a full worked answer key
10,000+ expert-curated prompts for ChatGPT, Claude, Gemini, and wherever you use AI. Our extension helps any prompt deliver better results.