ModelVerdict

Jiřina Švecová · 25401227

CzechClean image13 fields

Document cs-p0013 · layout cs-L06 · 13 fields · public part of the dataset · Same invoice in English (US) → · Report an error in this example

Jiřina Švecová · 25401227

Verdict

Cheapest without an error: DeepSeek V4.1 Flash at $0.42 per 1,000 documents.

  • GPT-6 Sol is also error-free but 34.7× more expensive.
  • Gemma 4 31B is 2.3× cheaper, but: wrong value (Supplier).

Prompt given to every model

Z přiloženého dokladu vytěž údaje a vrať pouze JSON s těmito klíči:

supplier_name      název dodavatele (přesně jak je na dokladu, včetně diakritiky a právní formy)
supplier_id        IČO dodavatele
vat_id             DIČ dodavatele
invoice_number     číslo dokladu
payment_reference  variabilní symbol
issue_date         datum vystavení (YYYY-MM-DD)
tax_point_date     datum uskutečnění zdanitelného plnění, DUZP (YYYY-MM-DD)
due_date           datum splatnosti (YYYY-MM-DD)
net_by_rate        základ daně podle sazeb DPH, např. {"21": 18450.00, "12": 1200.00}
vat_by_rate        DPH podle sazeb, např. {"21": 3874.50}
total              celkem k úhradě (číslo, záporné u dobropisu)
currency           měna podle ISO 4217 (CZK, EUR, …)
bank_account       číslo účtu dodavatele (tvar předčíslí-číslo/kód banky) nebo IBAN

Pravidla:
- Údaje odběratele nevracej, jen dodavatele.
- Pokud údaj na dokladu není, vrať null. Nic si nedomýšlej.
- Pokud doklad neobsahuje rozpis DPH (neplátce, zálohová faktura), vrať net_by_rate a vat_by_rate jako null.
- Částky jako čísla s desetinnou tečkou, bez oddělovačů tisíců a bez měny.

How we score

Exact match per field after normalising dates, amounts and whitespace. Missing diacritics count as an error. Cost is the real token spend of this run, converted at the rate shown in the footer.

What each model extracted

FieldCorrect valueGemma 4 31B11 / 13 fieldsDeepSeek V4.1 Flash13 / 13 fieldsMistral Small 412 / 13 fieldsGPT-6 Luna13 / 13 fieldsGemini 3.5 Flash Lite13 / 13 fieldsQwen3.8 Flash13 / 13 fieldsMistral Medium 3.510 / 13 fieldsGemini 3.8 Flash13 / 13 fieldsClaude Sonnet 513 / 13 fieldsGPT-6 Sol13 / 13 fields
SupplierJiřina Švecová Švřina Švecováwrong valueJiřina Švecová (correct)Jiřina Švecová (correct)Jiřina Švecová (correct)Jiřina Švecová (correct)Jiřina Švecová (correct)Jiřina Švecová (correct)Jiřina Švecová (correct)Jiřina Švecová (correct)Jiřina Švecová (correct)
Supplier ID (EIN, IČO…)1412175114121751 (correct)14121751 (correct)14121751 (correct)14121751 (correct)14121751 (correct)14121751 (correct) 12526100wrong value14121751 (correct)14121751 (correct)14121751 (correct)
VAT ID—— (correct)— (correct)— (correct)— (correct)— (correct)— (correct) CZ12526100invented value— (correct)— (correct)— (correct)
Invoice number2540122725401227 (correct)25401227 (correct)25401227 (correct)25401227 (correct)25401227 (correct)25401227 (correct)25401227 (correct)25401227 (correct)25401227 (correct)25401227 (correct)
Payment reference2540122725401227 (correct)25401227 (correct)25401227 (correct)25401227 (correct)25401227 (correct)25401227 (correct)25401227 (correct)25401227 (correct)25401227 (correct)25401227 (correct)
Invoice date2025-12-282025-12-28 (correct)2025-12-28 (correct)2025-12-28 (correct)2025-12-28 (correct)2025-12-28 (correct)2025-12-28 (correct)2025-12-28 (correct)2025-12-28 (correct)2025-12-28 (correct)2025-12-28 (correct)
Tax point—— (correct)— (correct) 2025-12-28invented value— (correct)— (correct)— (correct)— (correct)— (correct)— (correct)— (correct)
Due date2026-01-272026-01-27 (correct)2026-01-27 (correct)2026-01-27 (correct)2026-01-27 (correct)2026-01-27 (correct)2026-01-27 (correct)2026-01-27 (correct)2026-01-27 (correct)2026-01-27 (correct)2026-01-27 (correct)
Net by tax rate—— (correct)— (correct)— (correct)— (correct)— (correct)— (correct)— (correct)— (correct)— (correct)— (correct)
Tax by rate—— (correct)— (correct)— (correct)— (correct)— (correct)— (correct)— (correct)— (correct)— (correct)— (correct)
Total5729.825729.82 (correct)5729.82 (correct)5729.82 (correct)5729.82 (correct)5729.82 (correct)5729.82 (correct)5729.82 (correct)5729.82 (correct)5729.82 (correct)5729.82 (correct)
CurrencyCZKCZK (correct)CZK (correct)CZK (correct)CZK (correct)CZK (correct)CZK (correct)CZK (correct)CZK (correct)CZK (correct)CZK (correct)
Bank accountCZ0208000000942619491156 CZ02 0000 0000 9426 1949 1156wrong valueCZ02 0800 0000 9426 1949 1156 (correct)CZ02 0800 0000 9426 1949 1156 (correct)CZ02 0800 0000 9426 1949 1156 (correct)CZ02 0800 0000 9426 1949 1156 (correct)CZ02 0800 0000 9426 1949 1156 (correct) CZ02 0800 0000 009426 1949 1156wrong valueCZ02 0800 0000 9426 1949 1156 (correct)CZ02 0800 0000 9426 1949 1156 (correct)CZ02 0800 0000 9426 1949 1156 (correct)
Tokensinput / output704 / 2111,746 / 2572,592 / 1815,165 / 2021,516 / 2093,286 / 1,6282,592 / 1831,516 / 1,6125,317 / 2015,165 / 178
of which hidden reasoningbilled as output012308401,42701,403060
of which from cacheinput billed at a discount4276140006400000
Cost of this documentthis run$0.0002$0.0004$0.0005$0.0007$0.0010$0.0012$0.0053$0.0072$0.0126$0.0147
Per 1,000 documentsfor the same kind of document$0.18$0.42$0.50$0.75$0.98$1.17$5.26$7.18$12.64$14.69
Latencythis run7.7 s3.5 s1.7 s3.3 s2 s20.2 s3.2 s6.8 s3.8 s3.9 s