Properties
| Name | Type | Description |
|---|---|---|
| status | string | Always "success" here — errors come back as HTTP 4xx/5xx with the common error envelope instead. |
| data | object | |
| message | string |
JSON Schema
{
"$schema": "https://json-schema.org/draft/2020-12/schema",
"$id": "https://artifacts.apis.io/space-ocr-com/json-schema/space-ocr-com-ocr-fields-response-schema.json",
"title": "OcrFieldsResponse",
"x-generated": "2026-10-31",
"x-method": "derived",
"x-generator": "derive-json-schema.py",
"x-source": "openapi/space-ocr-com-openapi.json#/components/schemas/OcrFieldsResponse",
"type": "object",
"properties": {
"status": {
"type": "string",
"description": "Always \"success\" here — errors come back as HTTP 4xx/5xx with the common error envelope instead."
},
"data": {
"type": "object",
"properties": {
"values": {
"type": "object",
"additionalProperties": {},
"description": "Pure user data, exactly your declared schema — the model's reading of the page, held to the printed text by the character cross-check (not a byte-for-byte copy; for exact string matching use cells[path].evidence.printed_text)."
},
"normalized": {
"type": "object",
"additionalProperties": {},
"description": "Present only when some field declared a scalar type (number / integer / date, or a string carrying pattern / enum): a sparse tree shaped exactly like values, with the leaves parsed into the declared type. A leaf that would not parse is null — the reason is at cells[path].normalized.error. values itself is left alone; it is the side carrying the coordinates and the verification."
},
"cells": {
"type": "object",
"additionalProperties": {
"$ref": "#/$defs/Cell"
},
"description": "Flat, path-keyed coordinate/verification map — keys share review.flagged[].path grammar."
},
"review": {
"$ref": "#/$defs/Review"
},
"image": {
"$ref": "#/$defs/ImageInfo"
}
},
"required": [
"values",
"cells",
"review",
"image"
]
},
"message": {
"type": "string"
}
},
"required": [
"status",
"data"
],
"$defs": {
"Cell": {
"type": "object",
"properties": {
"box": {
"type": "object",
"properties": {
"xmin": {
"type": "number"
},
"ymin": {
"type": "number"
},
"xmax": {
"type": "number"
},
"ymax": {
"type": "number"
}
},
"required": [
"xmin",
"ymin",
"xmax",
"ymax"
],
"description": "Axis-aligned rectangle, normalized to 0–1000 — convert to pixels with data.image (pixel_x = xmin / 1000 × width)."
},
"quad": {
"type": "array",
"items": {
"type": "object",
"properties": {
"x": {
"type": "number"
},
"y": {
"type": "number"
}
},
"required": [
"x",
"y"
]
},
"description": "Four corners following a tilted scan — always present alongside box, same 0–1000 frame (the page as read; see data.image). No deskew is applied: coordinates stay in the input image's coordinate system."
},
"verified": {
"type": [
"boolean",
"null"
],
"description": "The cell's verdict — a mirror of review (engine v82): false whenever review carries reasons (any kind, declared rules included), true when nothing was flagged and a check actually ran, null when nothing was flagged but there was nothing to check (geometry-only units such as row unions and table containers). The character cross-check itself lives at evidence.text_match."
},
"review": {
"type": [
"object",
"null"
],
"properties": {
"reasons": {
"type": "array",
"items": {
"type": "string"
},
"description": "Every rule broken, ordered by rank — index 0 is the primary verdict (always an array, even at length one). Vocabulary: \"type_mismatch\" | \"out_of_range\" | \"pattern_mismatch\" | \"near_mismatch\" | \"near_ambiguous\" | \"near_conflict\" (the first six are rules the caller declared, ranked above engine-inferred reasons) | \"nobox\" | \"text_mismatch\" | \"crop_mismatch\" | \"low_ratio\" | \"weak_source\" | \"low_ocr_confidence\" | \"ambiguous_occurrence\" | \"overwide_box\" | \"missing\". New codes may be added later — fall back to a generic \"needs review\" for unknown ones."
}
},
"required": [
"reasons"
],
"description": "null = passed. Anything else = human review recommended; reasons carries the ranked verdicts."
},
"evidence": {
"$ref": "#/$defs/CellEvidence"
},
"normalized": {
"$ref": "#/$defs/CellNormalized"
}
},
"required": [
"box",
"quad",
"verified",
"review"
],
"description": "One entry of the flat, path-keyed coordinate/verification map. Keys share the grammar of review.flagged[].path (items[0].price); a row path like items[0] is the row's union box. Fields that came back without coordinates (nobox) or never came back (missing) have NO cell — they appear only in review.flagged."
},
"CellEvidence": {
"type": "object",
"properties": {
"text_match": {
"type": "boolean",
"description": "The character cross-check itself: did the value equal the OCR text under these coordinates (two independent engines agreeing)? Key present only when the check ran — when it could not run, verified is null. Engine v82 moved this out of `verified`, which is now the review verdict; `verified: false` with `text_match: true` is a normal combination (the characters matched and a declared rule caught it)."
},
"source": {
"type": "string",
"description": "Where the coordinates came from. Vocabulary varies by endpoint: vision_symbol_match / token_id / char_matcher_fallback / unclaimed_tokens / vision_paragraph …"
},
"match_ratio": {
"type": "number",
"description": "Match ratio of the character cross-check."
},
"printed_text": {
"type": "string",
"description": "What the OCR pass itself read at these coordinates (engine v83; /ocr/fields only). `values` is the model's rendering and the cross-check folds width / brackets / whitespace before comparing, so a value the model re-spelled still passes — use printed_text when you need an exact string match, such as reconciling against a master table. It does not supersede values (the OCR pass has its own misreads, which is why the two are cross-checked). Glyphs only — inter-word spacing is not reconstructed, so compare ignoring whitespace. Absent when nothing was matched (no box)."
},
"ocr_confidence": {
"type": "number",
"description": "Minimum OCR self-confidence over the matched glyphs. Key absent when unavailable. Not a review threshold and not declarable — gate on it caller-side if you need one."
},
"crop_verified": {
"type": "boolean",
"description": "Result of the crop re-read, present only when it ran."
},
"multiline": {
"type": "boolean",
"description": "Present (true) only when the value wraps and box is the union of those lines."
},
"near": {
"type": "object",
"properties": {
"occurrences": {
"type": "integer",
"description": "How many times this value is printed on the page."
},
"satisfied": {
"type": "integer",
"description": "How many of those sit beside a declared term."
},
"anchored": {
"type": "boolean",
"description": "Whether the copy that got the coordinates is one of them."
},
"nearest_term": {
"type": [
"string",
"null"
]
},
"nearest": {
"type": [
"number",
"null"
],
"description": "Gap to the closest declared term in multiples of the ALLOWED window — <= 1 would have passed (the window is anisotropic, 6 wide × 3 tall, so one absolute distance cannot be compared to one threshold)."
}
},
"required": [
"occurrences",
"satisfied",
"anchored"
],
"description": "Present only when a field that declared `near` was flagged near_mismatch / near_ambiguous — the arithmetic of the verdict, so a reviewer can re-derive it instead of trusting the reason."
},
"not_near": {
"type": "object",
"properties": {
"matched": {
"type": "string"
},
"distance": {
"type": "number"
}
},
"required": [
"matched",
"distance"
],
"description": "Present only on near_conflict — the declared term the value sat beside, distance in window multiples."
}
}
},
"CellNormalized": {
"type": "object",
"properties": {
"value": {
"description": "The parsed reading, or null when the printed text would not parse (see error)."
},
"type": {
"type": "string"
},
"method": {
"type": "string"
},
"error": {
"type": "string",
"description": "Present when value is null — why the printed text would not parse: \"not_numeric\" | \"not_an_integer\" | \"not_a_date\" | \"no_year\" | \"conventional_token\" (a legitimate printed convention such as 一式 — read correctly, just not the declared type) | \"relative_date\"."
}
},
"required": [
"type"
],
"description": "Deterministic second reading of the value in the declared scalar type. Present only on /ocr/fields cells whose field declared number / integer / date (or a string carrying pattern / enum)."
},
"FlaggedEntry": {
"type": "object",
"properties": {
"path": {
"type": "string",
"description": "Same grammar as the cells map keys — a direct lookup. Note: missing is raised only for fields declared required; an undeclared field that never comes back raises no flag."
},
"reasons": {
"type": "array",
"items": {
"type": "string"
},
"description": "Every rule broken, ordered by rank — index 0 is the primary verdict (always an array, even at length one). Vocabulary: \"type_mismatch\" | \"out_of_range\" | \"pattern_mismatch\" | \"near_mismatch\" | \"near_ambiguous\" | \"near_conflict\" (the first six are rules the caller declared, ranked above engine-inferred reasons) | \"nobox\" | \"text_mismatch\" | \"crop_mismatch\" | \"low_ratio\" | \"weak_source\" | \"low_ocr_confidence\" | \"ambiguous_occurrence\" | \"overwide_box\" | \"missing\". New codes may be added later — fall back to a generic \"needs review\" for unknown ones."
}
},
"required": [
"path",
"reasons"
]
},
"ImageInfo": {
"type": "object",
"properties": {
"width": {
"type": "integer"
},
"height": {
"type": "integer"
}
},
"required": [
"width",
"height"
],
"description": "The size in pixels of the page as read — the frame every coordinate lives in (pixel_x = xmin / 1000 × width). Measured after the page is turned upright (EXIF) and after any downscale, so it can differ from the file you sent: a sideways 4000×3000 photo can come back 3000×4000."
},
"Review": {
"type": "object",
"properties": {
"unit": {
"type": "string",
"description": "\"field\" on /ocr/fields, \"element\" on /ocr/markdown, \"block\" on /ocr/text."
},
"declared": {
"type": "integer",
"description": "/ocr/fields only — every declared slot, empty and unreturned required included (a fixed denominator)."
},
"returned": {
"type": "integer",
"description": "/ocr/fields only — non-empty values."
},
"total": {
"type": "integer",
"description": "/ocr/markdown and /ocr/text — total counted units (each table cell counts on its own)."
},
"boxed": {
"type": "integer",
"description": "Units that carry a box."
},
"verified": {
"type": "integer",
"description": "Units with verified: true — nothing flagged and a check actually ran. Flagged units are not counted here; their count is flagged.length."
},
"flagged": {
"type": "array",
"items": {
"$ref": "#/$defs/FlaggedEntry"
},
"description": "The review list — the review count IS its length (no separate counter)."
},
"by_reason": {
"type": "object",
"additionalProperties": {
"type": "integer"
},
"description": "Per-reason tally. One cell can break a declared rule AND carry an engine suspicion, so every entry in reasons is counted — the total is therefore at least flagged.length."
},
"coverage": {
"type": "object",
"properties": {
"recovered_blocks": {
"type": "number"
},
"vision_tokens": {
"type": "number"
},
"tokens_claimed": {
"type": "number"
},
"token_coverage": {
"type": "number"
}
},
"description": "How much of the page came back (document endpoints; LLM path only on /ocr/text)."
},
"notes": {
"type": "array",
"items": {
"$ref": "#/$defs/ReviewNote"
},
"description": "Present only when a declaration did not run as written — branch on issue."
}
},
"required": [
"unit",
"boxed",
"verified",
"flagged",
"by_reason"
]
},
"ReviewNote": {
"type": "object",
"properties": {
"path": {
"type": "string"
},
"issue": {
"type": "string",
"description": "What did not run as declared: \"type_coerced\" (unsupported declared type read as string; declared_type / applied_type ride along) | \"label_unresolved\" (a declared label anchored nothing — the value still comes back, so this note is the only trace that the declaration is inert) | \"near_unresolved\" (no declared near term is printed anywhere — the check abstained)."
},
"description": {
"type": "string"
}
},
"required": [
"path",
"issue"
]
}
}
}
Work with this as data
Every JSON Schema here is available over the APIs.io API and to AI agents over MCP.
MCP server
One button, every client — Claude, Cursor, VS Code and the rest.
https://apis.io/mcp
Tools for schemas
4 MCP tools reach this
find_json_schemasBrowse and filter every JSON Schema in the catalog.apis_io_searchSTART HERE — APIs, providers and tags for one query, each with its total.resolveTurn a domain, URL or GitHub org into the provider it belongs to.find_cohortsEvery scored population of providers in the catalog.
Call it yourself
curl for this page
This JSON Schema
curl "https://apis.io/api/v1/json-schemas/space-ocr-com-ocr-fields-response"
All schemas
curl "https://apis.io/api/v1/json-schemas?limit=25"
Discovery needs no key. Ratings and market analysis are Pro.
Get an API key
Free tier, no form to fill in. Signing in shares your email address with us — we store it to create your key and to recognise you if you sign in with another provider. See our Privacy Policy and Terms.
A second provider on the same verified email joins the account you already have.