> ## Documentation Index
> Fetch the complete documentation index at: https://composio-preview.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# PDF Vector

> PDF Vector parses documents into markdown, extracts structured data, answers document questions, and searches academic publications and grants.

![PDF Vector logo](https://mintcdn.com/composio-preview/gqjBKmmJCINafaiK/images/toolkits/pdf_vector.svg?fit=max\&auto=format\&n=gqjBKmmJCINafaiK\&q=85\&s=fa2ce7f53d0f1cbd1e6fb93bbcafc098)

* **Category:** ai document extraction
* **Auth:** API\_KEY
* **Composio-managed OAuth available?** N/A
* **Tools:** 10
* **Triggers:** 0
* **Slug:** `PDF_VECTOR`
* **Version:** 20261101\_00

## Tools

> [accordion] Ask Document
>
> **Slug:** `PDF_VECTOR_ASK_DOCUMENT`
>
> Answer one question from a general, identity, invoice, or bank-statement document. Processing consumes credits; the response reports the actual model and credit cost. The tool never retries with a higher model automatically.
>
> #### Input Parameters
>
> **file** `object`
>
> A document supplied through Composio's file input.
>
> **model** `string ("auto" | "nano" | "mini" | "pro" | "max")`
>
> Processing model. Higher models can consume more credits; this tool does not escalate automatically.
>
> **base64** `string`
>
> Base64-encoded document content. Prefer a URL or file for large documents; provide exactly one source.
>
> **question** `string` *required*
>
> Question to answer from the document (at least 4 characters).
>
> **document\_id** `string`
>
> Optional caller-defined document identifier echoed in the response.
>
> **document\_url** `string`
>
> Publicly fetchable URL of the document. Provide exactly one of document\_url, file, or base64.
>
> **document\_type** `string ("general" | "identity" | "invoice" | "bank_statement")`
>
> Document specialization to use. Choose identity, invoice, or bank\_statement only for that document class.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Extract Document Data
>
> **Slug:** `PDF_VECTOR_EXTRACT_DOCUMENT`
>
> Extract schema-constrained structured data from a general, identity, invoice, or bank-statement document. Processing consumes credits and never retries with a higher model automatically.
>
> #### Input Parameters
>
> **file** `object`
>
> A document supplied through Composio's file input.
>
> **model** `string ("auto" | "nano" | "mini" | "pro" | "max")`
>
> Processing model. Higher models can consume more credits; this tool does not escalate automatically.
>
> **base64** `string`
>
> Base64-encoded document content. Prefer a URL or file for large documents and provide exactly one source.
>
> **prompt** `string` *required*
>
> Extraction instructions describing the fields to return.
>
> **document\_id** `string`
>
> Optional caller-defined document identifier echoed in the response.
>
> **document\_url** `string`
>
> Publicly fetchable document URL. Provide exactly one of document\_url, file, or base64.
>
> **document\_type** `string ("general" | "identity" | "invoice" | "bank_statement")`
>
> Document specialization to use. Choose identity, invoice, or bank\_statement only for that document class.
>
> **output\_schema** `object` *required*
>
> JSON Schema that the extracted data must follow.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Fetch Publications
>
> **Slug:** `PDF_VECTOR_FETCH_PUBLICATIONS`
>
> Resolve up to 100 publication identifiers with provider auto-detection and return both resolved records and per-identifier errors.
>
> #### Input Parameters
>
> **ids** `array` *required*
>
> One to 100 DOI, PMID, arXiv, Semantic Scholar, ERIC, Europe PMC, or OpenAlex identifiers.
>
> **fields** `array`
>
> Publication fields to return; omit to use provider defaults. Choices: doi, title, url, providerURL, authors, date, year, totalCitations, totalReferences, abstract, pdfURL, provider, providerData.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Find Citations
>
> **Slug:** `PDF_VECTOR_FIND_CITATIONS`
>
> Find relevant academic citations for each sentence in a paragraph and preserve provider-specific partial errors.
>
> #### Input Parameters
>
> **fields** `array`
>
> Publication fields to include for each citation; omit for provider defaults. Choices: doi, title, url, providerURL, authors, date, year, totalCitations, totalReferences, abstract, pdfURL, provider, providerData.
>
> **paragraph** `string` *required*
>
> Paragraph whose sentences need supporting citations (1-5000 characters).
>
> **providers** `array`
>
> Academic databases to search for citations. Choices: semantic-scholar, pubmed, arxiv, google-scholar, eric, europe-pmc, openalex, crossref.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Find Similar Papers
>
> **Slug:** `PDF_VECTOR_FIND_SIMILAR_PAPERS`
>
> Find papers related to a seed publication using citation-network similarity, optionally including relationship edges.
>
> #### Input Parameters
>
> **id** `string` *required*
>
> Seed publication identifier (1-500 characters).
>
> **limit** `integer`
>
> Maximum similar papers to return (1-100).
>
> **fields** `array`
>
> Publication fields to return for the seed and similar papers. Choices: doi, title, url, providerURL, authors, date, year, totalCitations, totalReferences, abstract, pdfURL, provider, providerData.
>
> **include\_edges** `boolean`
>
> Include citing and cited-by identifier edges supporting each similarity result.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Get Paper Graph
>
> **Slug:** `PDF_VECTOR_GET_PAPER_GRAPH`
>
> Retrieve a publication plus one independently offset-controlled page of citing papers and one page of referenced papers.
>
> #### Input Parameters
>
> **id** `string` *required*
>
> Publication identifier for the graph root (1-500 characters).
>
> **fields** `array`
>
> Publication fields to return for the root, citing papers, and referenced papers; omit for provider defaults. Choices: doi, title, url, providerURL, authors, date, year, totalCitations, totalReferences, abstract, pdfURL, provider, providerData.
>
> **citations\_limit** `integer`
>
> Maximum citing papers to return (0-1000).
>
> **citations\_offset** `integer`
>
> Zero-based offset into citing papers.
>
> **references\_limit** `integer`
>
> Maximum referenced papers to return (0-1000).
>
> **references\_offset** `integer`
>
> Zero-based offset into referenced papers.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Parse Academic Paper
>
> **Slug:** `PDF_VECTOR_PARSE_ACADEMIC_PAPER`
>
> Resolve an academic identifier or provider URL to a public PDF and parse it to markdown. Processing consumes credits, returns the actual model and credit cost, and never retries with a higher model automatically.
>
> #### Input Parameters
>
> **model** `string ("auto" | "nano" | "mini" | "pro" | "max")`
>
> Parsing model. Higher models can consume more credits; the tool uses only the selected model and does not escalate automatically.
>
> **include\_pages** `boolean`
>
> Include markdown split into individual pages.
>
> **publication\_id** `string`
>
> DOI or supported provider identifier. Provide exactly one of publication\_id or publication\_url.
>
> **publication\_url** `string`
>
> Supported provider publication URL. Provide exactly one of publication\_url or publication\_id.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Parse Document
>
> **Slug:** `PDF_VECTOR_PARSE_DOCUMENT`
>
> Convert a general, identity, invoice, or bank-statement document to markdown; specialized documents require the auto, pro, or max model. Processing consumes credits; the response reports the actual model and credit cost, and the tool never retries with a higher model automatically.
>
> #### Input Parameters
>
> **file** `object`
>
> A document supplied through Composio's file input.
>
> **model** `string ("auto" | "nano" | "mini" | "pro" | "max")`
>
> Processing model. Higher models can consume more credits. Identity, invoice, and bank-statement parsing accept only auto, pro, or max.
>
> **base64** `string`
>
> Base64-encoded document content. Prefer a URL or file for large documents; provide exactly one source.
>
> **document\_id** `string`
>
> Optional caller-defined document identifier (1-512 characters) echoed in the response.
>
> **document\_url** `string`
>
> Publicly fetchable URL of the document. Provide exactly one of document\_url, file, or base64.
>
> **document\_type** `string ("general" | "identity" | "invoice" | "bank_statement")`
>
> Document specialization to use. Choose identity, invoice, or bank\_statement only for that document class.
>
> **include\_pages** `boolean`
>
> For general documents only, include markdown split into individual pages.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Search Publications
>
> **Slug:** `PDF_VECTOR_SEARCH_PUBLICATIONS`
>
> Search academic publications across selected databases and return one offset-controlled page, including any provider-specific partial errors.
>
> #### Input Parameters
>
> **limit** `integer`
>
> Maximum results per provider for this page (1-100).
>
> **query** `string` *required*
>
> Publication search query (1-400 characters).
>
> **fields** `array`
>
> Publication fields to return; omit for provider defaults. Choices: doi, title, url, providerURL, authors, date, year, totalCitations, totalReferences, abstract, pdfURL, provider, providerData.
>
> **offset** `integer`
>
> Zero-based result offset for this page; increase it by the number of returned results to request another page.
>
> **year\_to** `integer`
>
> Latest publication year, inclusive (1900-2100).
>
> **providers** `array`
>
> Academic databases to search. Choices: semantic-scholar, pubmed, arxiv, google-scholar, eric, europe-pmc, openalex, crossref.
>
> **year\_from** `integer`
>
> Earliest publication year, inclusive (1900-2100).
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Search Research Grants
>
> **Slug:** `PDF_VECTOR_SEARCH_RESEARCH_GRANTS`
>
> Search US, EU, and UK research-grant providers and return one explicit offset-controlled page with partial provider errors.
>
> #### Input Parameters
>
> **limit** `integer`
>
> Maximum results per provider for this page (1-50).
>
> **query** `string` *required*
>
> Grant search query (1-400 characters).
>
> **fields** `array`
>
> Grant fields to return; omit for provider defaults. Choices: sourceId, title, url, agency, program, description, eligibility, fundingAmountMin, fundingAmountMax, currency, deadlineDate, openDate, closeDate, grantType, region, keywords, piName, organizationName, provider, providerData.
>
> **offset** `integer`
>
> Zero-based result offset for this page; increase it by the number of returned results to request another page.
>
> **providers** `array`
>
> Grant databases to search; omit to search all supported providers. Choices: grants-gov, nih-reporter, cordis, ukri.
>
> **deadline\_to** `string`
>
> Latest grant deadline, inclusive, in YYYY-MM-DD format.
>
> **funding\_max** `number`
>
> Maximum funding amount; must be non-negative.
>
> **funding\_min** `number`
>
> Minimum funding amount; must be non-negative.
>
> **deadline\_from** `string`
>
> Earliest grant deadline, inclusive, in YYYY-MM-DD format.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not
