You're looking at a specific version of this model. Jump to the model overview.
Input schema
The fields you can use to run this model with an API. If you don’t give a value for a field its default value will be used.
| Field | Type | Default value | Description |
|---|---|---|---|
| image |
string
|
Screenshot or image file
|
|
| profile |
None
|
auto
|
Preset tuned for document type
|
| languages |
string
|
auto
|
Languages preset or custom list
|
| custom_languages |
string
|
|
Custom language codes (comma-separated) — used when languages=custom
|
| ocr_decoder |
None
|
beamsearch
|
Text decoder
|
| beam_width |
integer
|
7
Min: 1 Max: 15 |
Beam width (when decoder=beamsearch)
|
| allow_basic_punct |
boolean
|
True
|
Restrict to common punctuation to avoid odd symbols
|
| allow_brackets |
boolean
|
False
|
Permit brackets []{}<>
|
| merge_level |
None
|
auto
|
Text merging strength
|
| autotune |
boolean
|
True
|
Try a few parameter combos and pick best
|
| block_rare_symbols |
boolean
|
True
|
Block uncommon symbols ($§•◊ etc.)
|
| preprocessing |
boolean
|
True
|
Apply preprocessing (recommended)
|
| upscale_min_dim |
integer
|
1100
Max: 4000 |
Upscale small images so the shorter side reaches this size (px)
|
| clahe |
boolean
|
True
|
Apply mild CLAHE contrast enhancement
|
| denoise_strength |
integer
|
2
Max: 15 |
Denoising strength (0=off, 1-15)
|
| sharpen |
boolean
|
True
|
Apply mild unsharp mask to preserve small punctuation
|
| deskew |
boolean
|
True
|
Deskew small angles
|
| trim_borders |
boolean
|
True
|
Trim outer borders
|
| invert_strategy |
None
|
auto
|
Ensure dark text on light background
|
| binarize |
None
|
auto
|
Binarization mode
|
| morph_op |
None
|
none
|
Morphology op after binarize
|
| morph_kernel |
integer
|
0
Max: 15 |
Morph kernel size
|
| remove_shadow_strength |
integer
|
0
Max: 20 |
Uneven lighting removal (0=off)
|
| smart_preprocessing |
boolean
|
True
|
Auto-tune preprocessing based on image stats
|
| min_confidence |
number
|
0.3
Max: 1 |
Minimum confidence (0.0-1.0)
|
| text_only |
boolean
|
False
|
Return only text lines (list of strings)
|
| include_bboxes |
boolean
|
True
|
Include x1,y1,x2,y2 in output
|
| include_polygons |
boolean
|
False
|
Include 4-point polygon as flat list
|
| remove_page_numbers |
boolean
|
True
|
Remove likely page numbers near bottom
|
| dehyphenate |
boolean
|
True
|
Join hyphenated words across lines
|
| normalize_quotes |
boolean
|
True
|
Normalize common quotation marks and dashes
|
Output schema
The shape of the response you’ll get when you run this model with an API.
Schema
{'properties': {'markdown': {'format': 'uri',
'title': 'Markdown',
'type': 'string'},
'metadata': {'title': 'Metadata', 'type': 'string'}},
'required': ['markdown', 'metadata'],
'title': 'ModelOutput',
'type': 'object'}