> ## Documentation Index > Fetch the complete documentation index at: https://docs.querying.ai/llms.txt > Use this file to discover all available pages before exploring further. # Exporter les réponses notées > Exportez les réponses notées en JSON ou CSV, des plus récentes aux plus anciennes. ## OpenAPI ````yaml GET /v1/monitors/{id}/results openapi: 3.1.0 info: title: querying.ai description: >- One API for every AI answer engine. Every task is asynchronous: submit, then collect the result via webhook or polling. version: 1.0.0 servers: - url: https://api.querying.ai description: Production security: - bearerAuth: [] paths: /v1/monitors/{id}/results: get: tags: - Monitors summary: Export scored answers description: >- The scored rows themselves, newest first, for a warehouse load or a spreadsheet. since and until are half-open and both accept a bare YYYY-MM-DD (read as 00:00:00Z) for compatibility with the original export; every filter is applied server-side. Pagination is a keyset cursor over (ranAt, id) with microsecond precision, so equal timestamps cannot drop or repeat a row: keep the cursor's first response's since and until fixed and follow nextCursor until it is null. includeEvidence adds the retained answer text, the citations and the scoring context, and lowers the page size bound because a page of evidence is far heavier than a page of scores. operationId: listMonitorResults parameters: - name: id in: path required: true schema: type: string - name: since in: query required: false schema: type: string description: >- ISO instant or legacy YYYY-MM-DD (00:00:00Z). Default: 30 days ago. Inclusive. - name: until in: query required: false schema: type: string description: >- ISO instant or legacy YYYY-MM-DD (00:00:00Z). Default: now. Exclusive; must be after since. - name: engine in: query required: false schema: type: string - name: prompt in: query required: false schema: type: string maxLength: 2000 description: Exact prompt. - name: mentioned in: query required: false schema: type: boolean description: >- true keeps answers naming the brand; false keeps only the misses (the competitive-gap rows). - name: cited in: query required: false schema: type: boolean description: true keeps answers citing a registered domain. - name: competitor in: query required: false schema: type: string maxLength: 80 description: >- Keeps only answers whose stored competitorHits contain exactly this name (case-sensitive). It is matched against the stored hits, not checked against the monitor's current competitors, so a competitor removed since still filters its old answers and a misspelled name returns no rows rather than a 400. - name: includeEvidence in: query required: false schema: type: boolean default: false description: >- Add answerText, sources and scoringContext to every row, and lower the page-size bound to 100. - name: limit in: query required: false schema: type: integer minimum: 1 maximum: 10000 default: 10000 description: >- Default 10000 (max 10000) without evidence; default and maximum 100 when includeEvidence is true. - name: format in: query required: false schema: type: string enum: - json - csv default: json - name: cursor in: query required: false schema: type: string description: >- The previous response's nextCursor (JSON) or x-next-cursor header (CSV), unchanged. responses: '200': description: A page of scored rows, as JSON or CSV. content: application/json: schema: $ref: '#/components/schemas/MonitorResultsResponse' text/csv: schema: type: string example: >- ran_at,monitor,engine,prompt,mentioned,cited,position,competitors,task_id 2026-09-22T03:00:00.123456Z,Earbud brand tracking,CHATGPT,best wireless earbuds for commuting,true,false,412,Sony,0f4b6f1c-4c1e-4a4c-9d5a-2f0a5b0b1a2c headers: X-Next-Cursor: schema: type: string description: >- CSV only: the cursor for the next page, empty when the page is the last and nothing was truncated. X-Result-Truncated: schema: type: string enum: - 'true' - 'false' description: 'CSV only: whether more rows match than this page returned.' X-Result-Since: schema: type: string format: date-time description: 'CSV only: the resolved inclusive window start.' X-Result-Until: schema: type: string format: date-time description: 'CSV only: the resolved exclusive window end.' '400': $ref: '#/components/responses/ValidationError' '401': $ref: '#/components/responses/Unauthorized' '404': $ref: '#/components/responses/NotFound' components: schemas: MonitorResultsResponse: type: object properties: success: type: boolean const: true data: type: object properties: monitor: type: object properties: id: type: string name: type: string since: type: string format: date-time until: type: string format: date-time filters: type: object description: >- The filters that were applied, echoed back with null for the ones that were not. properties: engine: type: - string - 'null' prompt: type: - string - 'null' mentioned: type: - boolean - 'null' cited: type: - boolean - 'null' competitor: type: - string - 'null' truncated: type: boolean description: True when more rows matched than this page returned. nextCursor: type: - string - 'null' description: >- Pass back unchanged as cursor with the SAME since and until. Changing the window mid-pagination can skip or repeat rows. results: type: array items: $ref: '#/components/schemas/MonitorResultRow' MonitorResultRow: type: object description: >- One scored answer. position is the character offset of the earliest alias hit inside the answer text - a prominence proxy, not a rank, and null when the brand was absent (which position 0 would otherwise spell). properties: id: type: string format: uuid description: >- Scored-row id. The pagination cursor is built from it, so do not sort by it as if it were time. taskId: type: string engine: $ref: '#/components/schemas/MonitorEngine' prompt: type: string mentioned: type: boolean cited: type: boolean position: type: - integer - 'null' ranAt: type: string description: >- ISO instant with microsecond precision, in UTC. Used verbatim in the cursor. competitorHits: type: array items: type: string description: >- Names of this monitor's competitors that the answer mentioned, empty when none were. answerText: type: - string - 'null' description: >- Only with includeEvidence. The retained answer, markdown when the engine gave markdown, at most 8,000 characters with a trailing ellipsis when it was cut, and null when the engine returned no text. sources: type: array items: $ref: '#/components/schemas/MonitorStoredSource' description: Only with includeEvidence. scoringContext: $ref: '#/components/schemas/MonitorScoringContext' description: >- Only with includeEvidence. null on rows scored before it was recorded. ErrorResponse: type: object required: - success - error properties: success: type: boolean const: false error: $ref: '#/components/schemas/ErrorDetail' MonitorEngine: type: string description: >- An engine a monitor can schedule: the prompt surfaces sold to customers plus the deprecated NAVER alias. GOOGLE_AIO and GOOGLE_AIMODE are accepted as input aliases and echoed as GOOGLE and AIMODE. CITATION_ANALYSIS is an analysis task, not a prompt surface, and cannot be scheduled on a monitor. enum: - CHATGPT - GEMINI - PERPLEXITY - GOOGLE - AIMODE - NAVER_AI_BRIEF - NAVER_AI_TAB - NAVER MonitorStoredSource: type: object description: >- One citation as the engine gave it. The order is the engine's, and the list is deduplicated by URL. properties: url: type: string label: type: - string - 'null' position: type: integer description: >- 1-based place in the engine's citation list. It is not a ranking of your brand. MonitorScoringContext: type: - object - 'null' description: >- The definition this answer was scored with, stored at completion. null on rows scored before this was recorded - they are the reason a comparison can be withheld as legacy_scoring_unknown. properties: aliases: type: array items: type: string domains: type: array items: type: string competitors: type: array items: $ref: '#/components/schemas/MonitorCompetitor' country: type: string version: type: string description: >- Stable hash of the definition above. Two rows with the same version are comparable; a different one means the question changed. scoredAt: type: string format: date-time answerPresent: type: boolean description: False for a completed observation whose answer was empty. evidenceTruncated: type: boolean description: True when the retained answer text was cut at the character limit. ErrorDetail: type: object required: - code - message properties: code: type: string description: Stable, safe to branch on. See the Errors guide for the full list. message: type: string description: For humans. May change. details: description: >- Context for the error. An array of `{field, message}` on most `VALIDATION_ERROR`s; an object elsewhere (for example `{id}` on `RESOURCE_NOT_FOUND`, `retryAfterSec` on 429 and 503 codes). Omitted when there is nothing to add. oneOf: - type: array items: type: object properties: field: type: string message: type: string - type: object additionalProperties: true timestamp: type: string format: date-time MonitorCompetitor: type: object description: >- A brand scored alongside yours in the same answers. The name is always one of its own aliases, so a bare name matches reliably. required: - name properties: name: type: string maxLength: 80 description: __you__ is reserved for your own brand. aliases: type: array maxItems: 20 items: type: string description: Other spellings of the same brand. The name is added automatically. source: type: string enum: - user - auto readOnly: true description: >- user: you added it. auto: the weekly competitor read found it in the monitor's own answers. Entries you send with source auto are ignored, so a monitor read and written back keeps its found competitors. domain: type: string readOnly: true description: The competitor's site, when the competitor read found one. responses: ValidationError: description: >- The request is malformed — bad JSON, unrecognised `taskType`, or neither `payload.prompt` nor `payload.query` present. `details` names the offending field. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' examples: missingPrompt: summary: Neither prompt nor query supplied value: success: false error: code: VALIDATION_ERROR message: Request validation failed details: - field: payload.prompt message: payload.prompt or payload.query is required timestamp: '2026-07-09T04:12:00.000Z' badTaskType: summary: Unrecognised taskType value: success: false error: code: VALIDATION_ERROR message: Request validation failed details: - field: taskType message: '''NOPE'' is not recognized' timestamp: '2026-07-09T04:12:00.000Z' Unauthorized: description: Missing or invalid bearer token. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: success: false error: code: MISSING_API_KEY message: Missing or invalid API key timestamp: '2026-07-09T04:12:00.000Z' NotFound: description: >- No monitor with that id for this key, or no scored row with that task id. A monitor owned by another account is reported the same way rather than as a 403. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: success: false error: code: NOT_FOUND message: No monitor with that id timestamp: '2026-09-22T04:12:00.000Z' securitySchemes: bearerAuth: type: http scheme: bearer description: 'Your API key, sent as `Authorization: Bearer `.' ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.