Skip to main content

Common Mistakes Reference

Frequent Exa parameter-shape and deprecation mistakes, with corrections.
  • Base docs URL: https://exa.ai/docs
  • Search coding-agent reference: /reference/search-api-guide-for-coding-agents
  • Contents coding-agent reference: /reference/contents-api-guide-for-coding-agents
  • Monitors coding-agent reference: /reference/monitors-api-guide-for-coding-agents

Corrections

WrongCorrect
text: true at the top level on /searchNest it: "contents": {"text": true}
highlights: {...} at the top level on /searchNest it: "contents": {"highlights": {...}}
summary: true at the top level on /searchNest it: "contents": {"summary": true}
contents: { text: ... } on /contentsOn /contents, text, highlights, and summary are top-level fields
tokensNum on /search or /contentstokensNum belongs to /context, not search or contents
includeUrls / excludeUrlsUse includeDomains / excludeDomains
useAutoprompt in new requestsRemove it; it is deprecated
numSentences for highlightsUse maxCharacters or highlights: true
highlightsPerUrl for highlightsRemove it; it is deprecated
livecrawl as new default guidancePrefer maxAgeHours for new examples
livecrawl: "true"Do not pass the string "true"; it can silently fall back to never
Stacking text, highlights, and summary on every searchPick one. summary adds a per-result LLM call; combining text and highlights doubles billing for two views of the same page
category: "github", "documentation", "qa", "pdf"Stick to the documented category set. For code queries use type: "fast" or the /context endpoint
stream: true on /contents/contents does not support streaming
raw JSON field names copied into core Python SDK methods like search()Convert to snake_case
Python maxCharactersUse max_characters
Python outputSchemaUse output_schema
Python numResultsUse num_results
searchParams on monitorsUse search
schedule: "1h" on monitorsUse trigger: { "type": "interval", "period": "1h" }

High-Risk Shape Confusions

Search vs Contents

  • search endpoint: nested contents
  • contents endpoint: top-level text, highlights, summary
  • context endpoint: tokensNum
  • search endpoint: content sizing belongs under contents.text.maxCharacters
Last modified on June 26, 2026