> ## Documentation Index > Fetch the complete documentation index at: https://exa.ai/docs/llms.txt > Use this file to discover all available pages before exploring further. # Contents > Get the full page contents, summaries, and metadata for a list of URLs. Returns instant results from our cache, with automatic live crawling as fallback for uncached pages. *** ## OpenAPI ````yaml post /contents openapi: 3.1.0 info: version: 1.2.0 title: Exa Search API description: >- A comprehensive API for internet-scale search, allowing users to perform queries and retrieve results from a wide variety of sources using embeddings-based and traditional search. servers: - url: https://api.exa.ai security: - apikey: [] paths: /contents: post: summary: Get Contents operationId: getContents requestBody: required: true content: application/json: schema: type: object allOf: - type: object properties: urls: type: array description: >- Array of URLs to crawl (backwards compatible with 'ids' parameter). items: type: string example: - https://arxiv.org/pdf/2307.06435 default: - https://arxiv.org/pdf/2307.06435 ids: type: array description: Array of document IDs obtained from searches. items: type: string example: - https://arxiv.org/pdf/2307.06435 compliance: type: string enum: - hipaa description: >- Enterprise-only compliance mode. Set to `hipaa` for HIPAA mode. Uses cache-only retrieval; summaries and livecrawl are not supported. example: hipaa required: - urls - $ref: '#/components/schemas/ContentsRequest' responses: '200': $ref: '#/components/responses/ContentsResponse' x-codeSamples: - lang: bash label: Simple contents retrieval source: | curl -X POST 'https://api.exa.ai/contents' \ -H 'x-api-key: YOUR-EXA-API-KEY' \ -H 'Content-Type: application/json' \ -d '{ "urls": ["https://arxiv.org/abs/2307.06435"], "text": true }' - lang: python label: Simple contents retrieval source: | # pip install exa-py from exa_py import Exa exa = Exa('YOUR_EXA_API_KEY') results = exa.get_contents( urls=["https://arxiv.org/abs/2307.06435"], text=True ) print(results) - lang: javascript label: Simple contents retrieval source: | // npm install exa-js import Exa from 'exa-js'; const exa = new Exa('YOUR_EXA_API_KEY'); const results = await exa.getContents( ["https://arxiv.org/abs/2307.06435"], { text: true } ); console.log(results); - lang: php label: Simple contents retrieval source: '' - lang: go label: Simple contents retrieval source: '' - lang: java label: Simple contents retrieval source: '' - lang: bash label: Advanced contents retrieval source: | curl --request POST \ --url https://api.exa.ai/contents \ --header 'x-api-key: YOUR-EXA-API-KEY' \ --header 'Content-Type: application/json' \ --data '{ "urls": ["https://arxiv.org/abs/2307.06435"], "text": { "maxCharacters": 1000, "includeHtmlTags": false }, "highlights": { "query": "Key findings" }, "summary": { "query": "Main research contributions" }, "subpages": 1, "subpageTarget": "references", "extras": { "links": 2, "imageLinks": 1 } }' - lang: python label: Advanced contents retrieval source: | # pip install exa-py from exa_py import Exa exa = Exa('YOUR_EXA_API_KEY') results = exa.get_contents( urls=["https://arxiv.org/abs/2307.06435"], text={ "maxCharacters": 1000, "includeHtmlTags": False }, highlights={ "query": "Key findings" }, summary={ "query": "Main research contributions" }, subpages=1, subpage_target="references", extras={ "links": 2, "image_links": 1 } ) print(results) - lang: javascript label: Advanced contents retrieval source: | // npm install exa-js import Exa from 'exa-js'; const exa = new Exa('YOUR_EXA_API_KEY'); const results = await exa.getContents( ["https://arxiv.org/abs/2307.06435"], { text: { maxCharacters: 1000, includeHtmlTags: false }, highlights: { query: "Key findings" }, summary: { query: "Main research contributions" }, subpages: 1, subpageTarget: "references", extras: { links: 2, imageLinks: 1 } } ); console.log(results); components: schemas: ContentsRequest: type: object properties: text: oneOf: - type: boolean title: Simple text retrieval description: >- If true, returns full page text with default settings. If false, disables text return. - type: object title: Advanced text options description: >- Advanced options for controlling text extraction. Use this when you need to limit text length or include HTML structure. properties: maxCharacters: type: integer description: >- Maximum character limit for the full page text. Useful for controlling response size and API costs. example: 1000 includeHtmlTags: type: boolean default: false description: >- Include HTML tags in the response, which can help LLMs understand text structure and formatting. example: false verbosity: type: string enum: - compact - standard - full default: compact description: > Controls the verbosity level of returned content. Requires livecrawl: "always" to take effect. - compact: Most concise output, main content only (default) - standard: Balanced content with more detail - full: Complete content including all sections example: standard includeSections: type: array items: type: string enum: - header - navigation - banner - body - sidebar - footer - metadata description: > Only include content from these semantic page sections. Requires livecrawl: "always" to take effect. example: - body - header excludeSections: type: array items: type: string enum: - header - navigation - banner - body - sidebar - footer - metadata description: > Exclude content from these semantic page sections. Requires livecrawl: "always" to take effect. example: - navigation - footer - sidebar highlights: oneOf: - type: boolean title: Simple highlights retrieval description: >- If true, returns highlights with default settings. If false, disables highlights. - type: object title: Advanced highlights options description: >- Advanced options for steering highlight extraction. Pass `highlights: true` for the highest-quality default; supply this object only when you need to guide selection with your own query. properties: numSentences: type: integer minimum: 1 description: >- Deprecated and will be removed in a future release. Currently mapped to a character budget (1 sentence ≈ 1333 characters). Pass `highlights: true` for default highlights, or `{ query }` to guide selection with your own query. example: 1 deprecated: true highlightsPerUrl: type: integer minimum: 1 deprecated: true description: >- Deprecated and will be removed in a future release. Currently ignored. Pass `highlights: true` for default highlights, or `{ query }` to guide selection with your own query. example: 1 query: type: string description: Custom query that guides which highlights the LLM picks. example: Key advancements description: Text snippets the LLM identifies as most relevant from each page. properties: maxCharacters: type: integer minimum: 1 description: >- Maximum number of characters to return for highlights. Controls the total length of highlight text returned per URL. example: 2000 numSentences: type: integer minimum: 1 description: >- Deprecated and will be removed in a future release. Currently mapped to a character budget (1 sentence ≈ 1333 characters). Pass `highlights: true` for default highlights, or `{ query }` to guide selection with your own query. example: 1 deprecated: true highlightsPerUrl: type: integer minimum: 1 description: >- Deprecated and will be removed in a future release. Currently ignored. Pass `highlights: true` for default highlights, or `{ query }` to guide selection with your own query. deprecated: true query: type: string description: Custom query that guides which highlights the LLM picks. example: Key advancements summary: type: object description: Summary of the webpage properties: query: type: string description: Custom query for the LLM-generated summary. example: Main developments schema: type: object description: > JSON schema for structured output from summary. See https://json-schema.org/overview/what-is-jsonschema for JSON Schema documentation. example: $schema: http://json-schema.org/draft-07/schema# title: Title type: object properties: Property 1: type: string description: Description Property 2: type: string enum: - option 1 - option 2 - option 3 description: Description required: - Property 1 livecrawl: type: string enum: - never - fallback - preferred - always deprecated: true description: > **Deprecated**: Use `maxAgeHours` instead for more precise control over content freshness. Options for livecrawling pages. 'never': Disable livecrawling (default for neural search). 'fallback': Livecrawl when cache is empty. 'preferred': Always try to livecrawl, but fall back to cache if crawling fails. 'always': Always live-crawl, never use cache. Only use if you cannot tolerate any cached content. This option is not recommended unless consulted with the Exa team. example: preferred livecrawlTimeout: type: integer default: 10000 description: The timeout for livecrawling in milliseconds. example: 1000 maxAgeHours: type: integer description: > Maximum age of cached content in hours. Controls when livecrawling is triggered based on content freshness. - Positive value (e.g. 24): Use cached content if it's less than this many hours old, otherwise livecrawl. - 0: Always livecrawl, never use cache. - -1: Never livecrawl, always use cache. - Omit (default): Livecrawl as fallback only when no cached content exists. example: 24 subpages: type: integer default: 0 description: >- The number of subpages to crawl. The actual number crawled may be limited by system constraints. example: 1 subpageTarget: oneOf: - type: string - type: array items: type: string description: >- Term to find specific subpages of search results. Can be a single string or an array of strings, comma delimited. example: sources extras: type: object description: Extra parameters to pass. properties: links: type: integer default: 0 description: Number of URLs to return from each webpage. example: 1 imageLinks: type: integer default: 0 description: Number of images to return for each result. example: 1 context: oneOf: - type: boolean deprecated: true description: >- Deprecated: Use highlights or text instead. Returns page contents as a combined context string. example: true - type: object deprecated: true description: >- Deprecated: Use highlights or text instead. Returns page contents as a combined context string. properties: maxCharacters: type: integer description: Deprecated. Maximum character limit for the context string. example: 10000 ResultWithContent: allOf: - $ref: '#/components/schemas/Result' - type: object properties: text: type: string description: The full content text of the search result. example: >- Abstract Large Language Models (LLMs) have recently demonstrated remarkable capabilities... highlights: type: array items: type: string description: Array of highlights extracted from the search result content. example: - Such requirements have limited their adoption... highlightScores: type: array items: type: number format: float description: Array of cosine similarity scores for each highlighted example: - 0.4600165784358978 summary: type: string description: Summary of the webpage example: >- This overview paper on Large Language Models (LLMs) highlights key developments... subpages: type: array items: $ref: '#/components/schemas/ResultWithContent' description: Array of subpages for the search result. example: - id: https://arxiv.org/abs/2303.17580 url: https://arxiv.org/pdf/2303.17580.pdf title: >- HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face author: >- Yongliang Shen, Microsoft Research Asia, Kaitao Song, Microsoft Research Asia, Xu Tan, Microsoft Research Asia, Dongsheng Li, Microsoft Research Asia, Weiming Lu, Microsoft Research Asia, Yueting Zhuang, Microsoft Research Asia, yzhuang@zju.edu.cn, Zhejiang University, Microsoft Research Asia, Microsoft Research, Microsoft Research Asia publishedDate: '2023-11-16T01:36:20.486Z' text: >- HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face Date Published: 2023-05-25 Authors: Yongliang Shen, Microsoft Research Asia Kaitao Song, Microsoft Research Asia Xu Tan, Microsoft Research Asia Dongsheng Li, Microsoft Research Asia Weiming Lu, Microsoft Research Asia Yueting Zhuang, Microsoft Research Asia, yzhuang@zju.edu.cn Zhejiang University, Microsoft Research Asia Microsoft Research, Microsoft Research Asia Abstract Solving complicated AI tasks with different domains and modalities is a key step toward artificial general intelligence. While there are abundant AI models available for different domains and modalities, they cannot handle complicated AI tasks. Considering large language models (LLMs) have exhibited exceptional ability in language understanding, generation, interaction, and reasoning, we advocate that LLMs could act as a controller to manage existing AI models to solve complicated AI tasks and language could be a generic interface to empower t summary: >- HuggingGPT is a framework using ChatGPT as a central controller to orchestrate various AI models from Hugging Face to solve complex tasks. ChatGPT plans the task, selects appropriate models based on their descriptions, executes subtasks, and summarizes the results. This approach addresses limitations of LLMs by allowing them to handle multimodal data (vision, speech) and coordinate multiple models for complex tasks, paving the way for more advanced AI systems. highlights: - >- 2) Recently, some researchers started to investigate the integration of using tools or models in LLMs . highlightScores: - 0.32679107785224915 extras: type: object description: Results from extras. properties: links: type: array items: type: string description: Array of links from the search result. example: [] CostDollars: type: object properties: total: type: number format: float description: Total dollar cost for your request example: 0.007 breakDown: type: array description: Breakdown of costs by operation type items: type: object properties: search: type: number format: float description: Cost of your search operations example: 0.007 contents: type: number format: float description: Cost of your content operations example: 0 breakdown: type: object properties: neuralSearch: type: number format: float description: Cost of your neural search operations example: 0.007 deepSearch: type: number format: float description: Cost of your deep search operations example: 0.012 contentText: type: number format: float description: Cost of your text content retrieval example: 0 contentHighlight: type: number format: float description: Cost of your highlight generation example: 0 contentSummary: type: number format: float description: Cost of your summary generation example: 0 perRequestPrices: type: object description: Standard price per request for different operations properties: neuralSearch_1_10_results: type: number format: float description: >- Standard price for search with 1-10 results (contents for 10 results included) example: 0.007 neuralSearch_additional_result: type: number format: float description: Standard price per additional result beyond 10 example: 0.001 deepSearch: type: number format: float description: Standard price for deep search per request example: 0.012 deepReasoningSearch: type: number format: float description: Standard price for deep-reasoning search per request example: 0.015 perPagePrices: type: object description: Standard price per page for different content operations properties: contentText: type: number format: float description: >- Standard price per page for text content (included with search for first 10 results) example: 0.001 contentHighlight: type: number format: float description: >- Standard price per page for highlights (included with search for first 10 results) example: 0.001 contentSummary: type: number format: float description: Standard price per result for summaries example: 0.001 Result: type: object properties: title: type: string description: The title of the search result. example: A Comprehensive Overview of Large Language Models url: type: string format: uri description: The URL of the search result. example: https://arxiv.org/pdf/2307.06435.pdf publishedDate: type: - string - 'null' description: >- An estimate of the creation date, from parsing HTML content. Format is YYYY-MM-DD. example: '2023-11-16T01:36:32.547Z' author: type: - string - 'null' description: If available, the author of the content. example: >- Humza Naveed, University of Engineering and Technology (UET), Lahore, Pakistan id: type: string description: The temporary ID for the document. Useful for /contents endpoint. example: https://arxiv.org/abs/2307.06435 image: type: string format: uri description: The URL of an image associated with the search result, if available. example: https://arxiv.org/pdf/2307.06435.pdf/page_1.png favicon: type: string format: uri description: The URL of the favicon for the search result's domain. example: https://arxiv.org/favicon.ico responses: ContentsResponse: description: OK content: application/json: schema: type: object properties: requestId: type: string description: Unique identifier for the request example: e492118ccdedcba5088bfc4357a8a125 results: type: array items: $ref: '#/components/schemas/ResultWithContent' context: type: string description: >- Deprecated. Combined context string from search results. Use highlights or text instead. statuses: type: array description: Status information for each requested URL items: type: object properties: id: type: string description: The URL that was requested example: https://example.com status: type: string enum: - success - error description: Status of the content fetch operation example: success error: type: - object - 'null' description: Error details, only present when status is "error" properties: tag: type: string enum: - CRAWL_NOT_FOUND - CRAWL_TIMEOUT - CRAWL_LIVECRAWL_TIMEOUT - SOURCE_NOT_AVAILABLE - UNSUPPORTED_URL - CRAWL_UNKNOWN_ERROR description: Specific error type example: CRAWL_NOT_FOUND httpStatusCode: type: - integer - 'null' description: The corresponding HTTP status code example: 404 costDollars: $ref: '#/components/schemas/CostDollars' securitySchemes: apikey: type: apiKey name: x-api-key in: header description: >- API key can be provided either via x-api-key header or Authorization header with Bearer scheme ````