> ## Documentation Index > Fetch the complete documentation index at: https://docs.doconda.com/llms.txt > Use this file to discover all available pages before exploring further. # Read any file into Markdown ready for a model (Word, PowerPoint, Excel, PDF, .doc/.ppt/.xls, TXT, Markdown) ## OpenAPI ````yaml /openapi.json post /extract openapi: 3.1.0 info: title: Doconda API version: 0.1.0 license: name: Proprietary url: https://doconda.com/terms description: >- Create, review and edit Word, PowerPoint, Excel and PDF documents for AI agents. `POST /documents` works like an LLM API: direct (default), `stream: true` (SSE) or `background: true`. Errors follow RFC 9457 (application/problem+json). `POST /documents` accepts `Idempotency-Key`. servers: - url: https://api.eu.doconda.com/v1 description: EU (API keys `ak_eu_…`) - url: https://api.us.doconda.com/v1 description: US (not deployed yet) security: - apiKey: [] tags: - name: Documents description: Create, review and edit documents. - name: Files description: Your own files, to use as a source or to review/edit. - name: Webhooks description: Be notified when a document is ready. - name: Dashboard description: >- Used by the Doconda dashboard with a signed-in user. Not available with API keys. paths: /extract: post: tags: - Files summary: >- Read any file into Markdown ready for a model (Word, PowerPoint, Excel, PDF, .doc/.ppt/.xls, TXT, Markdown) operationId: extractFile requestBody: required: true content: application/json: schema: type: object properties: file: anyOf: - type: string pattern: ^file_ description: A file uploaded with POST /files. - type: string pattern: ^doc_ description: >- A document Doconda already made, by its id: we use its Word, PowerPoint or Excel (or its PDF, if that is all it has). - type: object properties: url: type: string maxLength: 2048 format: uri filename: type: string minLength: 1 maxLength: 200 required: - url additionalProperties: false description: We download it (public http/https, up to 20 MB). - type: object properties: data: type: string maxLength: 27962031 format: base64 contentEncoding: base64 pattern: >- ^$|^(?:[0-9a-zA-Z+/]{4})*(?:(?:[0-9a-zA-Z+/]{2}==)|(?:[0-9a-zA-Z+/]{3}=))?$ filename: type: string minLength: 1 maxLength: 200 required: - data - filename additionalProperties: false description: >- The file's bytes in base64, with its name (the extension says the format). store: default: true description: 'false: its content is deleted 15 minutes after it finishes.' type: boolean required: - file additionalProperties: false multipart/form-data: schema: type: object required: - file properties: file: type: string format: binary responses: '200': description: The text, as Markdown. content: application/json: schema: type: object properties: object: type: string const: extraction document_id: type: string pattern: ^doc_[0-9A-HJKMNP-TV-Z]{26}$ format: type: string enum: - docx - pptx - xlsx - pdf - doc - ppt - xls - txt - md - png - jpg - gif - webp markdown: type: string description: >- Headings, lists and tables kept; one section per slide or sheet; page markers in a PDF. Scanned pages and pictures are read with OCR. chars: type: integer minimum: -9007199254740991 maximum: 9007199254740991 description: Length of the whole text, before any cut. truncated: type: boolean description: true when longer than 2 000 000 characters. pages: anyOf: - type: integer minimum: -9007199254740991 maximum: 9007199254740991 - type: 'null' description: Pages (PDF), slides (PowerPoint) or sheets (Excel). required: - object - document_id - format - markdown - chars - truncated - pages '202': description: 'Past 120 s: follow the document with GET /documents/{id}.' content: application/json: schema: type: object properties: id: type: string pattern: ^doc_[0-9A-HJKMNP-TV-Z]{26}$ object: type: string const: document name: description: The name you gave it (`name` when creating it). type: - string - 'null' operation: type: string enum: - create - review - edit - extract format: anyOf: - type: string enum: - docx - pdf - pptx - xlsx - artifact description: >- The file you want. `docx` → Word (with a PDF too), `pdf` → only the PDF, `pptx` → PowerPoint (with a PDF), `xlsx` → Excel (with a PDF). `artifact` → a web page that keeps data (in each visitor's browser, or shared by everyone), on a signed link; with `store: false`, a page to download (see the artifacts guide). Without it, Doconda works it out from `prompt`. - type: 'null' type: anyOf: - type: string enum: - report - letter - contract - memo - minutes - presentation - spreadsheet - artifact description: >- What Doconda decided the document is. It sets the default look (a contract: serif, justified, numbered clauses; a letter: no page numbers). You don't choose it: say what you want in `prompt`, and the look in `style`. - type: 'null' mode: anyOf: - type: string enum: - prompt - content - manual - type: 'null' description: >- `manual`: a version someone edited by hand in the editor (POST /documents/{id}/versions). sector: description: >- The professional sector Doconda recognised in the request (legal, healthcare, accounting…). Its guide shapes the structure and conventions; what you say or send always wins. type: - string - 'null' parent_id: anyOf: - type: string pattern: ^doc_[0-9A-HJKMNP-TV-Z]{26}$ - type: 'null' description: The document this version was edited from. status: type: string enum: - queued - running - ready - ready_with_warnings - needs_review - failed - canceled phase: type: - string - 'null' pages: anyOf: - type: integer minimum: -9007199254740991 maximum: 9007199254740991 - type: 'null' style_applied: anyOf: - type: object properties: font: type: string font_size: type: number line_spacing: type: number colors: type: object properties: title: type: string headings: type: string text: type: string accent: type: string required: - title - headings - text - accent align: type: object properties: title: type: string enum: - left - center - right - justify headings: type: string enum: - left - center - right - justify body: type: string enum: - left - center - right - justify required: - title - headings - body margins_cm: type: number orientation: type: string enum: - portrait - landscape cover_page: type: boolean table_of_contents: type: boolean page_numbers: type: boolean header: type: string footer: type: string required: - font - font_size - line_spacing - colors - align - margins_cm - orientation - cover_page - table_of_contents - page_numbers - type: 'null' style_unsupported: type: array items: type: string validation_summary: anyOf: - type: object properties: errors: type: integer minimum: -9007199254740991 maximum: 9007199254740991 warnings: type: integer minimum: -9007199254740991 maximum: 9007199254740991 info: type: integer minimum: -9007199254740991 maximum: 9007199254740991 required: - errors - warnings - info - type: 'null' repair_summary: anyOf: - type: object properties: iterations: type: integer minimum: -9007199254740991 maximum: 9007199254740991 applied: type: integer minimum: -9007199254740991 maximum: 9007199254740991 rejected: type: integer minimum: -9007199254740991 maximum: 9007199254740991 required: - iterations - applied - rejected - type: 'null' error: anyOf: - type: object properties: code: type: string message: type: string required: - code - message - type: 'null' versions: type: object properties: engine: type: string renderer: type: string model: type: - string - 'null' metadata: type: object propertyNames: type: string additionalProperties: type: string store: type: boolean description: >- false: its content is deleted 15 minutes after it finishes. route: anyOf: - type: object properties: path: type: string enum: - fast - agent reason: type: string required: - path - reason description: >- The path Doconda chose to make the file, and why. `fast`: the AI writes the text and Doconda lays it out (seconds). `agent`: an AI agent writes the program that builds the file, looks at the pages and fixes them (30 s – 2 min; any design, comments, tracked changes, formulas and charts). Doconda always chooses. - type: 'null' quality: anyOf: - type: object properties: requested: type: string enum: - auto - fast - standard - best description: >- `auto`: Doconda picks the level the request needs (when in doubt, the cheaper one). level: anyOf: - type: string enum: - fast - standard - best description: >- How much work goes into the file. `fast`: one quick AI pass, seconds, the cheapest. `standard`: a careful pass and one look at the result, under a minute. `best`: an AI agent builds it like a designer would, looking at the pages and fixing them, a few minutes, the dearest. - type: 'null' description: The level used (null until Doconda decides it). reason: type: - string - 'null' required: - requested - level - reason - type: 'null' description: >- create and edit: the quality level asked for and the one used, which sets the price. artifact: anyOf: - type: object properties: storage: type: string enum: - local - shared description: >- Where an artifact's page keeps its data. `local`: in each visitor's browser (each one sees only theirs; nothing reaches Doconda). `shared`: in Doconda, the same for everyone with the link (polls, team lists). A project that stores nothing only allows `local`. required: - storage - type: 'null' description: 'Artifacts only: where the page keeps its data.' content_deleted_at: anyOf: - type: string format: date-time pattern: >- ^(?:(?:\d\d[2468][048]|\d\d[13579][26]|\d\d0[48]|[02468][048]00|[13579][26]00)-02-29|\d{4}-(?:(?:0[13578]|1[02])-(?:0[1-9]|[12]\d|3[01])|(?:0[469]|11)-(?:0[1-9]|[12]\d|30)|(?:02)-(?:0[1-9]|1\d|2[0-8])))T(?:(?:[01]\d|2[0-3]):[0-5]\d:[0-5]\d(?:\.\d+)?(?:Z|([+-](?:[01]\d|2[0-3]):[0-5]\d)))$ - type: 'null' description: >- When its files, request, report and events were deleted (`store: false`). outputs: description: Present once finished. type: array items: type: object properties: id: type: string pattern: ^out_[0-9A-HJKMNP-TV-Z]{26}$ kind: type: string enum: - docx - pptx - xlsx - pdf - preview - html - md page: type: integer minimum: -9007199254740991 maximum: 9007199254740991 content_type: type: string bytes: type: integer minimum: -9007199254740991 maximum: 9007199254740991 width: type: integer minimum: -9007199254740991 maximum: 9007199254740991 height: type: integer minimum: -9007199254740991 maximum: 9007199254740991 url: type: string description: >- Signed download URL, valid 5 minutes. For `html` (an artifact), the link to the page: 1 hour, or `?expires_in=` on GET /outputs; with `store: false`, the page to download. expires_at: type: string format: date-time pattern: >- ^(?:(?:\d\d[2468][048]|\d\d[13579][26]|\d\d0[48]|[02468][048]00|[13579][26]00)-02-29|\d{4}-(?:(?:0[13578]|1[02])-(?:0[1-9]|[12]\d|3[01])|(?:0[469]|11)-(?:0[1-9]|[12]\d|30)|(?:02)-(?:0[1-9]|1\d|2[0-8])))T(?:(?:[01]\d|2[0-3]):[0-5]\d:[0-5]\d(?:\.\d+)?(?:Z|([+-](?:[01]\d|2[0-3]):[0-5]\d)))$ required: - id - kind - content_type - bytes - url - expires_at created_at: type: string format: date-time pattern: >- ^(?:(?:\d\d[2468][048]|\d\d[13579][26]|\d\d0[48]|[02468][048]00|[13579][26]00)-02-29|\d{4}-(?:(?:0[13578]|1[02])-(?:0[1-9]|[12]\d|3[01])|(?:0[469]|11)-(?:0[1-9]|[12]\d|30)|(?:02)-(?:0[1-9]|1\d|2[0-8])))T(?:(?:[01]\d|2[0-3]):[0-5]\d:[0-5]\d(?:\.\d+)?(?:Z|([+-](?:[01]\d|2[0-3]):[0-5]\d)))$ completed_at: anyOf: - type: string format: date-time pattern: >- ^(?:(?:\d\d[2468][048]|\d\d[13579][26]|\d\d0[48]|[02468][048]00|[13579][26]00)-02-29|\d{4}-(?:(?:0[13578]|1[02])-(?:0[1-9]|[12]\d|3[01])|(?:0[469]|11)-(?:0[1-9]|[12]\d|30)|(?:02)-(?:0[1-9]|1\d|2[0-8])))T(?:(?:[01]\d|2[0-3]):[0-5]\d:[0-5]\d(?:\.\d+)?(?:Z|([+-](?:[01]\d|2[0-3]):[0-5]\d)))$ - type: 'null' required: - id - object - name - operation - format - type - mode - sector - parent_id - status - phase - pages - style_applied - style_unsupported - validation_summary - repair_summary - error - versions - metadata - store - route - quality - artifact - content_deleted_at - created_at - completed_at '401': description: >- unauthorized (missing, unknown or revoked key; expired session) or region_mismatch. content: application/problem+json: schema: type: object properties: type: type: string title: type: string status: type: integer minimum: -9007199254740991 maximum: 9007199254740991 code: type: string description: >- Stable machine code, e.g. input_invalid, insufficient_balance. request_id: type: string errors: type: array items: type: object properties: path: type: string message: type: string required: - path - message required: - type - title - status - code - request_id '413': description: file_too_large content: application/problem+json: schema: type: object properties: type: type: string title: type: string status: type: integer minimum: -9007199254740991 maximum: 9007199254740991 code: type: string description: >- Stable machine code, e.g. input_invalid, insufficient_balance. request_id: type: string errors: type: array items: type: object properties: path: type: string message: type: string required: - path - message required: - type - title - status - code - request_id '422': description: file_rejected / file_unreachable content: application/problem+json: schema: type: object properties: type: type: string title: type: string status: type: integer minimum: -9007199254740991 maximum: 9007199254740991 code: type: string description: >- Stable machine code, e.g. input_invalid, insufficient_balance. request_id: type: string errors: type: array items: type: object properties: path: type: string message: type: string required: - path - message required: - type - title - status - code - request_id security: - apiKey: [] - session: [] components: securitySchemes: apiKey: type: http scheme: bearer description: >- `Authorization: Bearer ak_eu_…`. The prefix is the project's region and must match the server (api.eu / api.us). session: type: http scheme: bearer bearerFormat: JWT description: 'Dashboard only: the Cognito ID token of the signed-in user.' ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.