> ## Documentation Index > Fetch the complete documentation index at: https://www.finta.com/docs/llms.txt > Use this file to discover all available pages before exploring further. # Retrieve export > Read an existing API export. Requires API access: Manage and Reports: View. Any authorized company teammate can retrieve it. Other companies, unknown IDs, raw database IDs, and exports created through the web UI return 404. Department exports also require the feature. GET never generates reports, dispatches workers, or marks a file downloaded. Status is pending, processing, succeeded, or failed. A file is exposed only after successful worker completion, final permission checks, and attachment. Fresh links are available for 30 days after completed_at. At or after that cutoff, status remains succeeded but file is null and error is null, including on an idempotent POST replay. The stored file is retained. Request a new export with a new Idempotency-Key to generate another report from the current books. Within that window, signed download URLs request expiry seven days after retrieval; retrieve this resource again for a fresh URL. Previously issued links keep their original expiry, which can extend beyond the 30-day regeneration cutoff. Signing credential expiry or storage policies can invalidate a link sooner. Link expiry does not delete the file. A download started before expiry can finish afterward; restarting or requesting further byte ranges after expiry requires a fresh link. Download URLs are bearer links and remain usable until expiration even if API access changes. Both responses use Cache-Control no-store. Failure details contain safe codes, never stack traces. The existing API request log stores export ID, status, and whether a link was issued, alongside caller and request time. Pending/failed polls and HEAD probes do not count as link issuance. Actual S3 downloads are not tracked. started_at records the first generation attempt and can remain null when access is revoked before generation; generation_duration_ms is the cumulative time spent in generation attempts, excluding queue and retry waits. ## OpenAPI ````yaml /openapi.yaml get /reports/exports/{id} openapi: 3.0.3 info: title: Finta API version: '1.0' description: > The Finta API provides programmatic access to your company's financial data, including transactions, journal entries, categories, financial reports, and the team directory. ## Authentication Each API key is scoped to a single company. There is no company ID parameter. The key already knows which company it belongs to. If you have access to multiple companies, create a separate key for each. A key acts as its owner; it does not have a separate permission snapshot. The owner must currently have **API access: Manage**, plus the permission required by the endpoint: | Endpoint | Required permission | |----------|---------------------| | `GET /company` | Company: View | | `GET /taxes`, `GET /taxes/{id}` | Taxes: View | | `POST /files`, `POST /file_uploads`, `POST /taxes/income_uploads`, `POST /taxes/franchise_uploads` | Taxes: Manage | | `GET /team` | Team & access: View | | `DELETE /team/invitations/{id}` | Team & access: Manage; target-management limits apply. No additional Growth/RBAC gate. | | `POST /team/invitations/{id}/resend` | Team & access: Manage; target-management limits apply. No additional Growth/RBAC gate. | | `PATCH /team/invitations/{id}` | Team & access: Manage; RBAC and active/trialing Growth required, with grant and target limits. | | `POST /team/invitations` | Team & access: Manage; requested grants must also be allowed for the person and plan. | | `GET /categories`, `GET /departments`, `GET /rules`, `GET /rules/{id}` | Accounting setup: View | | `POST /rules` | Accounting setup: Manage; applying to existing transactions (`apply_to_existing_transactions`, default true) also requires Transactions: Manage | | `PATCH /rules/{id}` | Accounting setup: Manage; Transactions: Manage is not required | | `DELETE /rules/{id}` | Accounting setup: Manage and Transactions: Manage, including absent retries | | `GET /integrations` | Integrations: View | | `GET /tax_credits`, `GET /tax_credits/{id}` | Taxes: View | | `GET /reports/*`, `POST /reports/exports` | Reports: View | | `GET /accounts`, `GET /cards`, `GET /transactions`, `GET /transactions/{id}` | Transactions: View | | `GET /forms/1099`, `GET /forms/1099/{id}` | Taxes: View | | `POST /transactions/bulk/update`, `GET /jobs/{id}` | Transactions: Manage | | `PATCH /transactions/{id}`, `POST /accounts`, `POST /transaction_uploads`, `POST /transfers`, `DELETE /transfers/{id}` | Transactions: Manage | | `GET /journal_entries`, `GET /journal_entries/{id}`, `GET /aggregations/total`, `GET /aggregations/totals`, `GET /parties/merchants` | Reports: View | | `GET /invoices`, `GET /invoices/{id}` | Invoices: View | | `POST /invoices`, `POST /invoices/{id}/void` | Invoices: Manage | | `POST /invoices/{id}/match`, `POST /invoices/{id}/unmatch` | Invoices: Manage and Transactions: Manage | Permission changes apply to the next request. Removing API access permanently revokes the owner's active keys for that company; restoring access requires creating a new key. Authenticate by including your API key in the `Authorization` header: ``` Authorization: Bearer finta_... ``` ### Key prefix convention Every Finta credential is prefixed so it can be recognized at a glance in logs, error messages, and secret scanners. Today there is one credential type and its prefix is `finta_`. As additional credential or token types are introduced (for example restricted keys or webhook signing secrets), each will have its own distinct prefix that stacks on the brand prefix (e.g. `finta__...`). Treat the prefix as load-bearing: do not strip it before sending, and do not assume a missing or unknown prefix is still a Finta credential. ## Rate Limits The API enforces two independent limits. Both surface as `429 Too Many Requests`, but they are distinct error codes with different retry guidance. ### Per-minute burst limit Authenticated requests are limited to **6,000 requests per 60 seconds** (100 per second) per API key. Unauthenticated requests (missing or invalid key) are limited to **120 requests per 60 seconds** per IP address. The window is **fixed**, not rolling. Each 60-second wall-clock interval is its own counter; all requests within that interval share it, and the counter resets to 0 in a single step at the next boundary. Available budget jumps from `0` back to the full bucket size (`6000` for authenticated API keys, `120` for unauthenticated IP buckets) instantaneously at the boundary, rather than aging out one slot at a time. Trust `X-RateLimit-Reset` for the exact reset timestamp. Successful responses and burst-limit rejections include the following headers. Do not assume they are present on other errors: early authentication, authorization, and monthly-limit rejections can omit them. This describes existing header behavior, not a new quota policy. | Header | Description | |--------|-------------| | `X-RateLimit-Limit` | Maximum requests per window (`6000` for authenticated requests, `120` for unauthenticated). | | `X-RateLimit-Remaining` | Requests remaining in the current window. | | `X-RateLimit-Reset` | Unix timestamp when the window resets. | On a 429 for this case, an additional header is included: | Header | Description | |--------|-------------| | `Retry-After` | Number of seconds until the next request is safe. | The error body is `type: rate_limit_error` / `code: rate_limit_exceeded`. **Retry guidance:** This case is retryable. Honor `Retry-After` as the minimum wait, then apply exponential backoff with jitter for any retries beyond the first (cap at a few minutes; abandon after a small number of attempts to avoid retry storms). Do not retry tighter than `Retry-After`. ### Monthly per-company limit Each company has a monthly cap on total API calls, independent of the per-minute burst: | Plan | Monthly cap | |------|-------------| | Formation | 10 | | Startup | 300 | | Growth | 30,000 | | Enterprise | Unlimited | Enterprise is enabled internally by Finta for contract-managed companies. It bypasses commercial monthly API/MCP quotas and ordinary subscription lockouts, even when the Stripe subscription is absent or canceled. Requests and MCP tool calls are still counted. Enterprise does not bypass valid credentials, company/user permissions, company closure, request validation, payload bounds, or per-minute abuse limits. Disabling Enterprise restores ordinary plan gates using the current billing state; it does not change Stripe billing. Counts reset at **midnight Pacific Time (US & Canada)** on the 1st of each month. This is the Finta application's configured time zone; the cap is computed in that zone, not in UTC. Consumers in other time zones should convert the local-midnight Pacific boundary to their own clock; note that the absolute UTC offset shifts by one hour twice a year for daylight saving (PST is UTC-8, PDT is UTC-7). Blocked requests are not counted toward the limit. The error body is `type: limit_error` / `code: monthly_limit_exceeded`, and the `message` field includes the reset date in human-readable form. **Retry guidance:** This case is **not retryable in the short term.** No `Retry-After` header is sent, deliberately, because the only meaningful remediation is to upgrade the plan (Settings -> Plans at app.finta.com) or wait until the 1st of the next month. Naive retry loops should branch on `error.code` and stop retrying when they see `monthly_limit_exceeded`. SDKs that auto-retry on 429 should look at `error.code` (or the absence of `Retry-After`) before re-issuing. A temporary company limit override, set by Finta support for the current month, may change the allowance and the wording of the monthly-limit message. It applies to every user and API key in the company. The error code and no-short-term-retry rule remain unchanged. ## Pagination List endpoints use cursor-based pagination. Pass `limit` to control page size (default 100, max 500). To fetch the next page, pass `starting_after` with the `next_cursor` value from the previous response. Responses include `object: "list"`, `url` (the canonical path of the collection, relative to the API host), `has_more` (whether more results exist), `next_cursor` (an opaque value to pass unchanged as `starting_after`), and `data` (the page of items). ## Reporting dates and freshness Reports and category totals use calendar dates in **America/Los_Angeles (Pacific Time)**, including defaults such as today and the current quarter. Periodic requests include both `start_date` and `end_date` exactly. Point-in-time requests include activity through the exact `date`. Response dates identify the boundaries actually used, not rounded months. Whole months read whole-month rollups. The current month's first day through today in Pacific Time reads month-to-date (MTD) rollups when fresh. Other partial-month boundaries, including a midmonth start, read raw journal entries for just those partial months and are slower; any whole months in the same request still use rollups. Stale or unfilled MTD rollups safely fall back to raw entries. Whole-month requests and month-aligned month-start-to-today or year-to-date requests stay on the fast rollup path when MTD is fresh. Fresh MTD must end at exactly the requested cutoff. If a refresh advances the rollup across Pacific midnight during a request, exact live entries preserve the original cutoff; later-day MTD is never included in that answer. `refreshed_at` is a Unix timestamp covering the entire report, total, or grouped page, including partial months. For rollup-only reads it is the maximum contributing rollup timestamp, including MTD. Any live raw-entry portion is current as of the call, even when no raw entries match. A fresh MTD row contributes its timestamp even when it contains zero activity through today; future-only merchants do not appear as empty merchant groups. A grouped page covers its returned groups; a later month's refresh does not advance an earlier monthly page's timestamp. It is null only when no contributing timestamp or live raw read is available. Rollup-backed activity may lag newly imported or edited entries. ## Amounts Every monetary field in the API follows two non-negotiable rules: 1. **Integer cents.** Monetary values are integers, never floats and never formatted strings. `$499.00` is `49900`. `$1,500.00` is `150000`. Divide by 100 to get dollars. This avoids floating-point precision errors in financial calculations. 2. **`_cents` suffix.** The field name always ends in `_cents` (e.g. `amount_cents`, `balance_cents`, `total_cents`, `net_income_cents`, `change_in_cash_cents`, `cash_cents`). If a field name does not end in `_cents`, it is not a monetary value. These rules apply to every monetary field across every endpoint (transactions, journal entries, income statement, balance sheet, cash flow). There are no exceptions and there is no "summary" or "display" sibling field that returns the same amount in dollars or as a string. If you encounter a field that appears to hold a monetary value but does not end in `_cents`, or whose value is not an integer, treat it as a bug and report it. Sign conventions: - **Income statement**: revenue/income amounts are positive, expense amounts are negative. Summing every section's `total_cents` yields net income without sign-flipping. - **Balance sheet**: `balance_cents` carries the natural-sign cumulative balance for that account at the report date. - **Cash flow**: `amount_cents` is positive for cash inflows and negative for cash outflows. ## Errors Errors return a consistent JSON structure: ```json { "error": { "type": "invalid_request_error", "code": "resource_missing", "message": "The requested resource was not found.", "doc_url": "https://www.finta.com/docs/api-reference/errors#resource_missing" } } ``` ### Status codes | Status | Meaning | |--------|---------| | `200 OK` | Request succeeded. | | `201 Created` | A new resource was created, such as an invitation, an account, or a transaction upload draft. | | `400 Bad Request` | Request was malformed or had invalid parameters. | | `401 Unauthorized` | The bearer credential is missing, malformed, or unrecognized. The fix is to obtain a valid API key. | | `403 Forbidden` | The API key is valid, but the caller does not have permission to access the resource. The fix is to regain the required company or product-area access. See the `Forbidden` response component for the specific codes. | | `404 Not Found` | The requested resource, or the endpoint itself, does not exist. Unknown or renamed paths (for example an old hyphenated report path like `/reports/income-statement` instead of `/reports/income_statement`) return this same envelope with `type: invalid_request_error` and `code: resource_missing`, as JSON, regardless of the `Accept` header. | | `422 Unprocessable Entity` | Invitation data failed validation, for example an invalid email or a conflicting invitation or existing teammate. | | `429 Too Many Requests` | Either the per-minute burst rate limit (retryable, with `Retry-After`) or the monthly per-company API call limit (not retryable in the short term, no `Retry-After`). Branch on `error.code` to distinguish the two and avoid retry storms on monthly exhaustion. See the Rate Limits section for the full retry guidance. | | `5xx` | An unexpected error on Finta's side. Retry with exponential backoff. | The `error.type` and `error.code` fields are stable identifiers safe to switch on programmatically. The `error.message` is human-readable and may change without notice. ### Early access endpoints Some endpoints are in early access: documented here, but turned on per company while they roll out. For a company that doesn't have one yet, every request to it returns `403 Forbidden` with `type: permission_error` and `code: feature_disabled`, ahead of the endpoint's own permission check, and changes nothing: ```json { "error": { "type": "permission_error", "code": "feature_disabled", "message": "This feature isn’t turned on for this company yet. Contact Finta to request access.", "doc_url": "https://www.finta.com/docs/api-reference/errors#feature_disabled" } } ``` It is not retryable: contact Finta to request access for the company, then retry. Once the feature is on for the company the endpoint works normally, with no change to requests. servers: - url: https://app.finta.com/api/v1 description: Production security: - bearerAuth: [] tags: - name: Company description: Retrieve company metadata - name: Plans & billing description: >- The plans Finta offers, what the company pays, and how much of each plan limit it has used. - name: Team description: Retrieve company members and manage invitations - name: Categories description: List your chart of accounts categories - name: Rules description: Company automation rule definitions. - name: Departments description: List the departments available to the authenticated company - name: Integrations description: List the authenticated company's integrations and their connection status - name: Aggregations description: >- Compute totals for one category and its descendants, as a single figure or grouped by merchant or month - name: Transactions description: > Transactions are what your bank shows you: a single line like "AWS charged you $499." Most API consumers want transactions. For accounting-level detail, see Journal Entries. Amounts are in cents (e.g. `amount_cents: 49900` = $499.00). - name: Journal Entries description: > Journal entries are the accounting records that transactions create. A single $499 AWS charge becomes two journal entries: debit Software $499, credit Cash $499 (double-entry bookkeeping). Use journal entries when you need accounting-level detail. Amounts are in cents (e.g. `amount_cents: 49900` = $499.00). - name: Invoices description: > Invoices the company sent, and the bank deposits that paid them. Recording an invoice books its revenue; Finta doesn't send invoices to customers. Amounts are in cents (e.g. `amount_cents: 49900` = $499.00). - name: Parties description: Discover observed counterparties for category-scoped financial queries. - name: Forms description: Discover vendor compliance filing records and yearly progress. - name: Reports description: > Financial statements: income statement, balance sheet, and cash flow. Each report returns the complete statement for a single period. You cannot filter to one category or request multiple periods in one call. - name: Feedback description: Send questions, feedback and bug reports to Finta support - name: Tax Credits description: Credit amounts and claim progress by type, jurisdiction, and tax year - name: Taxes description: >- Discover income-tax and Delaware franchise-tax filing records and available PDFs. - name: Files description: Upload the PDFs that tax uploads attach to the company's tax records. paths: /reports/exports/{id}: get: tags: - Reports summary: Retrieve export description: > Read an existing API export. Requires API access: Manage and Reports: View. Any authorized company teammate can retrieve it. Other companies, unknown IDs, raw database IDs, and exports created through the web UI return 404. Department exports also require the feature. GET never generates reports, dispatches workers, or marks a file downloaded. Status is pending, processing, succeeded, or failed. A file is exposed only after successful worker completion, final permission checks, and attachment. Fresh links are available for 30 days after completed_at. At or after that cutoff, status remains succeeded but file is null and error is null, including on an idempotent POST replay. The stored file is retained. Request a new export with a new Idempotency-Key to generate another report from the current books. Within that window, signed download URLs request expiry seven days after retrieval; retrieve this resource again for a fresh URL. Previously issued links keep their original expiry, which can extend beyond the 30-day regeneration cutoff. Signing credential expiry or storage policies can invalidate a link sooner. Link expiry does not delete the file. A download started before expiry can finish afterward; restarting or requesting further byte ranges after expiry requires a fresh link. Download URLs are bearer links and remain usable until expiration even if API access changes. Both responses use Cache-Control no-store. Failure details contain safe codes, never stack traces. The existing API request log stores export ID, status, and whether a link was issued, alongside caller and request time. Pending/failed polls and HEAD probes do not count as link issuance. Actual S3 downloads are not tracked. started_at records the first generation attempt and can remain null when access is revoked before generation; generation_duration_ms is the cumulative time spent in generation attempts, excluding queue and retry waits. operationId: retrieveExport parameters: - name: id in: path required: true schema: type: string description: Export ID from creation, starting with export_. responses: '200': description: Current export status and result. content: application/json: schema: $ref: '#/components/schemas/ReportExport' '401': description: Authentication required. content: application/json: schema: $ref: '#/components/schemas/Error' '403': description: Report/API permission or department feature unavailable. content: application/json: schema: $ref: '#/components/schemas/Error' '404': description: No API export found for this ID in this company. content: application/json: schema: $ref: '#/components/schemas/Error' '429': description: API rate or monthly limit exceeded. content: application/json: schema: $ref: '#/components/schemas/Error' components: schemas: ReportExport: type: object required: - id - object - report_type - start_date - end_date - format - group_by - status - created - started_at - completed_at - generation_duration_ms - file - error properties: id: type: string pattern: >- ^export_[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}$ object: type: string enum: - report_export report_type: type: string enum: - income_statement - balance_sheet - cash_flow format: type: string enum: - pdf - csv - xlsx start_date: type: string format: date end_date: type: string format: date group_by: allOf: - $ref: '#/components/schemas/ReportExportGrouping' - required: - period - rows status: type: string enum: - pending - processing - succeeded - failed description: >- pending awaits generation; processing includes generation and retry waits; succeeded exposes the retained file after final access checks while within 30 days of completed_at, then file is null; failed is terminal and requires a new key for a new job. An HTTP 200 retrieval may still carry failed status. created: type: integer started_at: type: integer nullable: true description: >- Unix timestamp of the first generation attempt. Null until generation starts, including a terminal failure when access is revoked before generation. completed_at: type: integer nullable: true description: >- Unix timestamp of terminal completion, whether succeeded or failed. Null while unfinished. generation_duration_ms: type: integer nullable: true minimum: 0 description: >- Cumulative generation attempt duration in milliseconds, including upload but excluding queue and retry wait time. Null until an attempt finishes. file: type: object nullable: true description: >- Available only for succeeded exports before completed_at plus 30 days. Null at or after that cutoff, including on idempotent POST replay; status remains succeeded and error remains null. Within the window, refresh download_url by retrieving the export again. Already-issued links keep their original expiry; neither link expiry nor the regeneration cutoff removes the stored file. required: - filename - content_type - size - download_url - expires_at properties: filename: type: string content_type: type: string enum: - application/pdf - text/csv - >- application/vnd.openxmlformats-officedocument.spreadsheetml.sheet size: type: integer minimum: 0 download_url: type: string format: uri expires_at: type: integer description: >- Requested Unix timestamp of download link expiry, seven days after retrieval. Signing credentials or storage policies may invalidate it sooner. This does not expire the stored file. error: type: object nullable: true description: >- Safe job failure details, present only when status is failed. This is distinct from an HTTP error envelope. New submissions with unconfirmed dispatch return HTTP 503 and retain a pending job, rather than setting enqueue_failed. required: - code - message properties: code: type: string enum: - access_revoked - enqueue_failed - generation_failed message: type: string Error: type: object required: - error properties: error: type: object required: - type - code - message - doc_url properties: type: type: string enum: - authentication_error - permission_error - invalid_request_error - rate_limit_error - limit_error - api_error description: >- The error category. `authentication_error` (401) means the bearer credential was missing, malformed, or unrecognized. `permission_error` (403) means the credential was valid but the caller is not allowed to access the resource. `invalid_request_error` covers invalid parameters, unavailable resources, invitation validation failures, and requests that conflict with current state (409). `rate_limit_error` is for per-minute burst limits (retry shortly). `limit_error` is for monthly API call limits per company (upgrade or wait for reset). `api_error` denotes an operation failure on the server (500), or temporarily unavailable rule definitions (503 rules_not_ready). Retry with exponential backoff; persistent 503 errors require operator attention. code: type: string enum: - invalid_api_key - user_inactive - user_removed_from_company - company_closed - subscription_required - staging_access_denied - insufficient_permission - resource_missing - parameter_missing - category_update_failed - transaction_update_failed - rate_limit_exceeded - monthly_limit_exceeded - invalid_date_range - invalid_date - invalid_request - idempotency_key_conflict - plan_upgrade_required - feature_disabled - invitation_invalid - invitation_create_failed - permissions_update_failed - invitation_resend_failed - invitation_revoke_failed - integration_not_connectable - integration_already_connected - connect_link_in_progress - connect_link_failed - rules_not_ready - forms_not_ready - taxes_not_ready - rule_already_exists - rule_busy - rule_deletion_failed - account_create_failed - upload_committed - upload_commit_failed - feedback_unavailable - closed_books - invoice_already_exists - invoice_not_open - invoice_not_matched - payment_already_matched - transaction_not_matchable - invoice_operation_failed - transaction_already_matched - transfer_not_eligible - transfer_operation_failed - report_export_failed - report_export_limit_exceeded - invalid_file - wrong_document_type - file_already_attached - tax_upload_not_allowed - tax_upload_failed description: A specific error code. message: type: string description: A human-readable description of the error. doc_url: type: string format: uri description: A link to documentation about this error. example: https://www.finta.com/docs/api-reference/errors#invalid_api_key param: type: string nullable: true description: The parameter that caused the error, if applicable. ReportExportGrouping: type: object additionalProperties: false description: >- Omitted members default to month for period and category for rows. Department rows are supported only for income statements and require the feature. properties: period: type: string enum: - month - quarter - year default: month description: >- Time columns by month, quarter, or year. Quarter/year boundaries follow the company's calendar/fiscal report settings. First and last columns may cover partial periods. rows: type: string enum: - category - department default: category description: >- category uses accounting categories. department uses the GAAP department income statement; it is supported only for income_statement and requires the departments feature. securitySchemes: bearerAuth: type: http scheme: bearer description: API key prefixed with `finta_` ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.