specification: FinOps Framework
specificationVersion: '1.0'
schema: https://www.finops.org/framework/
provider: Mathpix
providerId: mathpix
created: '2026-05-25'
modified: '2026-05-25'
reconciled: true
tags:
  - FinOps
  - Cost Management
  - FOCUS
  - OCR
  - Document Conversion
description: FOCUS-aligned FinOps definition for the Mathpix Convert API. Usage-based metering by
  image, page, and conversion. Apps are aggregated under a single account billing surface, with
  per-app_id and per-usage_type breakdowns available through the OCR Usage API.
sources:
  - https://mathpix.com/pricing
  - https://docs.mathpix.com
  - https://docs.mathpix.com/reference/ocr-usage
alignedWith:
  framework: FinOps Foundation Framework
  frameworkUrl: https://www.finops.org/framework/
  dataSpec: FOCUS
  dataSpecVersion: '1.3'
  dataSpecUrl: https://focus.finops.org/focus-specification/v1-3/
publisherName: Mathpix
serviceCategory: AI and Machine Learning
billingModel:
  pricingCategory: Usage-Based
  billingFrequency: Monthly
  billingCurrency: USD
  chargeCategories:
    - Usage
    - Purchase
    - Tax
    - Credit
    - Adjustment
  primaryUnit: image
  secondaryUnits:
    - page
    - conversion
    - request
focusColumns:
  ServiceName: Mathpix Convert API
  ServiceCategory: AI and Machine Learning
  ProviderName: Mathpix
  PublisherName: Mathpix
  InvoiceIssuerName: Mathpix
  PricingCategory: Usage-Based
  PricingUnit: image
  BillingCurrency: USD
meters:
  - name: image_ocr_requests
    unit: image
    aggregation: sum
    dimensions:
      - app_id
      - usage_type
      - request_args_hash
    source:
      api: /v3/ocr-usage
      group_by: usage_type
      timespan: hour
  - name: document_ocr_pages
    unit: page
    aggregation: sum
    dimensions:
      - app_id
      - usage_type
    source:
      api: /v3/ocr-usage
      group_by: usage_type
      timespan: day
  - name: conversion_requests
    unit: conversion
    aggregation: sum
    dimensions:
      - app_id
      - format
    source:
      api: /v3/ocr-usage
      group_by: usage_type
      timespan: day
  - name: strokes_requests
    unit: request
    aggregation: sum
    dimensions:
      - app_id
  - name: batch_images
    unit: image
    aggregation: sum
    dimensions:
      - app_id
      - batch_id
allocation:
  primaryKey: app_id
  secondaryKeys:
    - usage_type
    - request_args_hash
  notes: Mathpix recommends issuing distinct app_ids per workload (e.g., snip-web, ingestion-pipeline,
    notebook-grader) so cost can be allocated cleanly via the OCR Usage API's group_by=app_id mode.
forecasting:
  recommendedSignals:
    - monthly_image_ocr_count
    - monthly_document_pages
    - p95_pages_per_pdf
    - batch_throughput
  notes: For high-volume forecasts, combine /v3/ocr-usage with batch-level metadata captured at
    submission time. The Snip Pro $4.99/mo subscription is forecast as a fixed per-user charge.
optimization:
  recommendations:
    - Batch related images via /v3/batch to reduce per-request overhead.
    - Use page_ranges on /v3/pdf to avoid OCRing pages your workflow does not need.
    - Cache conversion_id outputs and avoid re-issuing /v3/converter for unchanged MMD inputs.
    - Mint short-lived app_tokens for client-side capture instead of embedding app_key in client code.