Invoices and forms
Insurance claim checkbox correction 2
Original source · Answer-key source
CC0-1.0. Made-up fixture generated for The Hype Kevin; dedicated to CC0. No real entities or personal data.
The answer key
Check numbers below match the grid columns. An absence check requires the stated value to be absent in its stated field.
| Check | Kind | Expected answer | Field / row / column |
|---|---|---|---|
| 1 | text | CL-FABLE-411 | |
| 2 | text | Fictional Moss Comet Studio | |
| 3 | blank | Purchase order (required) |
Model-by-check results
Right and Wrong describe exact checks. No output scores zero. Tab reaches each result.
| Model | Page checks right | Page score / 100 | 1 | 2 | 3 |
|---|---|---|---|---|---|
| Claude Opus 5.5 | 3/3 | 100.0 | Right | Right | Right |
| GPT-6.1 Sol | 3/3 | 100.0 | Right | Right | Right |
| Claude Sonnet 5.5 | 3/3 | 100.0 | Right | Right | Right |
| Gemini 3.8 Flash | 3/3 | 100.0 | Right | Right | Right |
| Gemini 4 Argon | 3/3 | 100.0 | Right | Right | Right |
| Chandra OCR 2 4-bit | 3/3 | 100.0 | Right | Right | Right |
| Mistral OCR | 3/3 | 100.0 | Right | Right | Right |
| GPT-6 Luna | 3/3 | 100.0 | Right | Right | Right |
| LightOnOCR 3 1B | 2/3 | 66.7 | Right | Right | Wrong |
| OCRia-1 | 0/3 | 0.0 | Wrong | Wrong | Wrong |
| dots.ocr 4-bit | 3/3 | 100.0 | Right | Right | Right |
| Nemotron 3.5 Parse | 3/3 | 100.0 | Right | Right | Right |
| Indra | 3/3 | 100.0 | Right | Right | Right |
| Granite-Docling | 3/3 | 100.0 | Right | Right | Right |
| PaddleOCR-VL 1.6 | 3/3 | 100.0 | Right | Right | Right |
| Tesseract 5.5.3 | 3/3 | 100.0 | Right | Right | Right |
Raw model outputs beside the key
Key reference
CL-FABLE-411Fictional Moss Comet Studio
Claude Opus 5.5 · 3/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
FICTIONAL SAMPLE • All entities and addresses are invented • CC0 # Fictional Umbrella Pebble Insurance ## Claim CL-FABLE-411 Claimant: Fictional Moss Comet Studio Date: 2026-10-22 Purchase order (required): blank --- Other cover: N/A --- Loss amount USD: 2,608.45 --- Delivery: ☑ Standard ~~☑~~ ☒ Express (ticked box crossed out — cancelled) --- *Crossed box cancelled; new tick applies. QZ* Contact: ☑ Email ☐ Phone --- Inspection: ☑ Yes ☐ No --- checkbox-form-2 · checkbox-form-2 · page 1
GPT-6.1 Sol · 3/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
FICTIONAL SAMPLE • All entities and addresses are invented • CC0 # Fictional Umbrella Pebble Insurance ## Claim CL-FABLE-411 Claimant: Fictional Moss Comet Studio Date: 2026-10-22 Purchase order (required): blank --- Other cover: N/A --- Loss amount USD: 2,608.45 --- Delivery: ☑ Standard ~~☑~~ Express --- Crossed box cancelled; new tick applies. QZ Contact: ☑ Email ☐ Phone --- Inspection: ☑ Yes ☐ No --- checkbox-form-2 · checkbox-form-2 · page 1
Claude Sonnet 5.5 · 3/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
FICTIONAL SAMPLE • All entities and addresses are invented • CC0 # Fictional Umbrella Pebble Insurance ## Claim CL-FABLE-411 Claimant: Fictional Moss Comet Studio Date: 2026-10-22 Purchase order (required): blank --- Other cover: N/A --- Loss amount USD: 2,608.45 --- Delivery: ☑ Standard ~~☒ Express~~ --- *Crossed box cancelled; new tick applies. QZ* Contact: ☑ Email ☐ Phone --- Inspection: ☑ Yes ☐ No --- checkbox-form-2 · checkbox-form-2 · page 1
Gemini 3.8 Flash · 3/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
FICTIONAL SAMPLE • All entities and addresses are invented • CC0 # Fictional Umbrella Pebble Insurance ### Claim CL-FABLE-411 Claimant: Fictional Moss Comet Studio Date: 2026-10-22 Purchase order (required): blank --- Other cover: N/A --- Loss amount USD: 2,608.45 --- Delivery: [x] Standard [~~x~~] Express --- Crossed box cancelled; new tick applies. QZ Contact: [x] Email [ ] Phone --- Inspection: [x] Yes [ ] No --- checkbox-form-2 · checkbox-form-2 · page 1
Gemini 4 Argon · 3/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
FICTIONAL SAMPLE • All entities and addresses are invented • CC0 # Fictional Umbrella Pebble Insurance ## Claim CL-FABLE-411 Claimant: Fictional Moss Comet Studio Date: 2026-10-22 Purchase order (required): blank --- Other cover: N/A --- Loss amount USD: 2,608.45 --- Delivery: [x] Standard ~~[x] Express~~ --- Crossed box cancelled; new tick applies. QZ Contact: [x] Email [ ] Phone --- Inspection: [x] Yes [ ] No --- checkbox-form-2 · checkbox-form-2 · page 1
Chandra OCR 2 · 3/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
<div data-bbox="53 38 454 52" data-label="Page-Header"><p>FICTIONAL SAMPLE • All entities and addresses are invented • CC0</p></div><div data-bbox="53 63 705 93" data-label="Section-Header"><h1>Fictional Umbrella Pebble Insurance</h1></div><div data-bbox="53 108 314 130" data-label="Section-Header"><h2>Claim CL-FABLE-411</h2></div><div data-bbox="53 145 395 162" data-label="Text"><p>Claimant: Fictional Moss Comet Studio</p></div><div data-bbox="53 166 211 183" data-label="Text"><p>Date: 2026-10-22</p></div><div data-bbox="53 211 508 230" data-label="Text"><p>Purchase order (required): _____</p></div><div data-bbox="53 257 204 274" data-label="Text"><p>Other cover: N/A</p></div><div data-bbox="53 303 304 321" data-label="Text"><p>Loss amount USD: 2,608.45</p></div><div data-bbox="53 350 373 372" data-label="Text"><p>Delivery: <input checked="" type="checkbox"/> Standard <input checked="" type="checkbox"/> Express</p></div><div data-bbox="53 400 548 424" data-label="Text"><p><i>Crossed box cancelled; new tick applies. @Z</i></p></div><div data-bbox="53 452 326 473" data-label="Text"><p>Contact: <input checked="" type="checkbox"/> Email <input type="checkbox"/> Phone</p></div><div data-bbox="53 503 298 523" data-label="Text"><p>Inspection: <input checked="" type="checkbox"/> Yes <input type="checkbox"/> No</p></div><div data-bbox="53 960 314 975" data-label="Page-Footer"><p>checkbox-form-2 · checkbox-form-2 · page 1</p></div>
Mistral OCR · 3/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
<file name="checkbox-form-2.pdf"> FICTIONAL SAMPLE • All entities and addresses are invented • CC0 # Fictional Umbrella Pebble Insurance ## Claim CL-FABLE-411 Claimant: Fictional Moss Comet Studio Date: 2026-10-22 Purchase order (required): _________________________ Other cover: N/A Loss amount USD: 2,608.45 Delivery: ☑ Standard ☒ Express *Crossed box cancelled; new tick applies. QZ* Contact: ☑ Email ☐ Phone Inspection: ☑ Yes ☐ No checkbox-form-2 · checkbox-form-2 · page 1 </file>
GPT-6 Luna · 3/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
FICTIONAL SAMPLE • All entities and addresses are invented • CC0 # Fictional Umbrella Pebble Insurance ## Claim CL-FABLE-411 Claimant: Fictional Moss Comet Studio Date: 2026-10-22 Purchase order (required): blank Other cover: N/A Loss amount USD: 2,608.45 Delivery: [✓] Standard [crossed-out box] Express Crossed box cancelled; new tick applies. QZ Contact: [✓] Email [ ] Phone Inspection: [✓] Yes [ ] No checkbox-form-2 · checkbox-form-2 · page 1
LightOnOCR 3 1B · 2/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
 FICTIONAL SAMPLE • All entities and addresses are invented • CC0  # Fictional Umbrella Pebble Insurance  ## Claim CL-FABLE-411  Claimant: Fictional Moss Comet Studio Date: 2026-10-22  Purchase order (required): _________________________  Other cover: N/A  Loss amount USD: 2,608.45  Delivery: ☑ Standard ✗ Express  *Crossed box cancelled; new tick applies. QZ*  Contact: ☑ Email □ Phone  Inspection: ☑ Yes □ No  checkbox-form-2 · checkbox-form-2 · page 1
OCRia-1 · 0/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
checkbox-form-2 · checkbox-form-2 · page 1
dots.ocr · 3/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
FICTIONAL SAMPLE • All entities and addresses are invented • CC0 # Fictional Umbrella Pebble Insurance Claim CL-FABLE-411 Claimant: Fictional Moss Comet Studio Date: 2026-10-22 Purchase order (required): _________________ Other cover: N/A Loss amount USD: 2,608.45 Delivery: ☑ Standard ☑ Express Crossed box cancelled; new tick applies. QZ Contact: ☑ Email ☐ Phone Inspection: ☑ Yes ☐ No
Nemotron 3.5 Parse · 3/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
FICTIONAL SAMPLE • All entities and addresses are invented • CC0 # **Fictional Umbrella Pebble Insurance** ## **Claim CL-FABLE-411** Claimant: Fictional Moss Comet Studio Date: 2026-10-22 Purchase order (required): __________________ Other cover: N/A Loss amount USD: 2,608.45 Delivery: ☑ Standard ☒ Express Crossed box cancelled; new tick applies. QZ Contact: ☑ Email ☐ Phone Inspection: ☑ Yes ☐ No checkbox-form-2 · checkbox-form-2 · page 1
Indra · 3/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
## Fictional Umbrella Pebble Insurance Claim CL-FABLE-411 Claimant: Fictional Moss Comet Studio Date: 2026-10-22 Purchase order (required): Other cover: N/A Loss amount USD: 2,608.45 Delivery: [V] Standard ] Express Crossed box cancelled; new tick applies. 2Z Contact: [] Email ] Phone Inspection: [] Yes No
Granite-Docling · 3/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
## Fictional Umbrella Pebble Insurance ## Claim CL-FABLE-411 Claimant: Fictional Moss Comet Studio Date: 2026-10-22 Purchase order (required): \_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_ Other cover: N/A Loss amount USD: 2,608.45 Delivery: Standard Express Crossed box cancelled; new tick applies. Qz Contact: Email Phone Inspection: Yes No
PaddleOCR-VL 1.6 · 3/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
FICTIONAL SAMPLE • All entities and addresses are invented • CC0 Fictional Umbrella Pebble Insurance Claim CL-FABLE-411 Claimant: Fictional Moss Comet Studio Date: 2026-10-22 Purchase order (required): _____ Other cover: N/A Loss amount USD: 2,608.45 Delivery: ☑ Standard ☒ Express Crossed box cancelled; new tick applies. QZ Contact: ☑ Email ☐ Phone Inspection: ☑ Yes ☐ No checkbox-form-2 · checkbox-form-2 · page 1
Tesseract 5.5.3 · 3/3 checks in scoring unit
Scoring unit: checkbox-form-2. Joined pages use the same full output.
FICTIONAL SAMPLE * All entities and addresses are invented + CCO Fictional Umbrella Pebble Insurance Claim CL-FABLE-411 Claimant: Fictional Moss Comet Studio Date: 2026-10-22 Purchase order (required): Other cover: N/A Loss amount USD: 2,608.45 Delivery: Standard x Express Crossed box cancelled; new tick applies. QZ, Contact: Email [_] Phone Inspection: [Y] Yes [_] No checkbox-form-2 - checkbox-form-2 - page 1