Cruz Logistics Inc. · Seattle Plumbing Co.
Contract text (the model got exactly this text)
MASTER SERVICES AGREEMENT This Master Services Agreement (the “Agreement”) is entered into by and between Cruz Logistics Inc., a Delaware corporation with its principal office at 3464 Technology Drive, Suite 326, Pittsburgh, PA 15276, EIN 93-9701112 (“Provider”) and Seattle Plumbing Co., a limited liability company with its principal office at 2457 Pine Street, Suite 132, Seattle, WA 98133, EIN None (“Client”). 1. SCOPE 1.1 Provider shall provide Client with IT infrastructure management and end-user support services as described in Exhibit A. 2. FEES AND PAYMENT 2.1 Client shall pay Provider $21,250 as a one-time fee, exclusive of applicable sales tax ($22,737.50 including 7% sales tax). 2.2 Invoices are payable within thirty (30) days of the invoice date. 2.3 Within ten (10) days of signing, Client shall pay an advance payment of $4,250. 2.4 Any amount not paid when due shall accrue a late charge of 0.01% of the overdue amount per day until paid. 3. TERM AND TERMINATION 3.1 This Agreement shall become effective on the first day of the calendar month following the date of its signature. 3.2 The term of this Agreement ends on December 31, 2028. 3.3 Either party may terminate this Agreement for convenience upon three (3) months' prior written notice to the other party. 4. CONFIDENTIALITY 4.1 Each party shall keep the other party's Confidential Information strictly confidential. 5. MISCELLANEOUS 5.1 This Agreement shall be governed by and construed in accordance with the laws of the State of New York, without regard to its conflict of laws principles. 5.2 This Agreement constitutes the entire agreement between the parties. Signed in New Heidi on April 11, 2025. Provider: ____________________ Client: ____________________
Verdict
Cheapest without an error: Gemma 4 31B at $0.19 per 1,000 documents.
- Claude Sonnet 5 is also error-free but 29.1× more expensive.
- Mistral Small 4 is 1.1× cheaper, but: wrong value (Provider).
Prompt given to every model
Extract the key terms of the contract below and return only JSON with these keys:
provider_name the party that delivers (provider, seller, landlord, licensor, contractor),
exactly as written, including the legal form
provider_id that party's company ID (EIN), if stated
client_name the party that pays (client, buyer, tenant, licensee, customer)
client_id that party's company ID (EIN), if stated
contract_type one of: services, supply, lease, license, maintenance
price the agreed price excluding VAT / sales tax (number)
currency ISO 4217 code (USD, EUR, CZK, …)
price_period one of: one_time, monthly, yearly
effective_date the date the contract takes effect (YYYY-MM-DD)
end_date the date a fixed term ends (YYYY-MM-DD); null if the term is indefinite
notice_period_months the termination notice period in months (number); null if there is none
penalty_pct_per_day the late-payment penalty in percent of the overdue amount per day
(e.g. 0.05 for 0.05 %); null if there is none
governing_law ISO 3166 country code of the governing law; US states as US-XX (e.g. US-NY)
Rules:
- Resolve relative dates ("on the date of signature", "the first day of the following month")
against the signing date.
- The signing date is not the effective date unless the contract says so.
- Deposits, advance payments and fixed penalties for other breaches are not the price or the
late-payment penalty.
- If a value is not in the contract, return null. Do not guess.
- Amounts as numbers with a decimal point, no thousands separators and no currency symbol.
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
| Field | Correct value | Mistral Small 411 / 13 fields | Gemma 4 31B13 / 13 fields | GPT-6 Luna13 / 13 fields | DeepSeek V4.1 Flash13 / 13 fields | Qwen3.8 Flash11 / 13 fields | Gemini 3.5 Flash Lite13 / 13 fields | Mistral Medium 3.513 / 13 fields | Gemini 3.8 Flash13 / 13 fields | GPT-6 Sol13 / 13 fields | Claude Sonnet 513 / 13 fields |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Provider | Cruz Logistics Inc. | Cruz Logistics Inc., a Delaware corporationwrong value | Cruz Logistics Inc. (correct) | Cruz Logistics Inc. (correct) | Cruz Logistics Inc. (correct) | Cruz Logistics Inc., a Delaware corporationwrong value | Cruz Logistics Inc. (correct) | Cruz Logistics Inc. (correct) | Cruz Logistics Inc. (correct) | Cruz Logistics Inc. (correct) | Cruz Logistics Inc. (correct) |
| Provider ID | 93-9701112 | 93-9701112 (correct) | 93-9701112 (correct) | 93-9701112 (correct) | 93-9701112 (correct) | 93-9701112 (correct) | 93-9701112 (correct) | 93-9701112 (correct) | 93-9701112 (correct) | 93-9701112 (correct) | 93-9701112 (correct) |
| Client | Seattle Plumbing Co. | Seattle Plumbing Co., a limited liability companywrong value | Seattle Plumbing Co. (correct) | Seattle Plumbing Co. (correct) | Seattle Plumbing Co. (correct) | Seattle Plumbing Co., a limited liability companywrong value | Seattle Plumbing Co. (correct) | Seattle Plumbing Co. (correct) | Seattle Plumbing Co. (correct) | Seattle Plumbing Co. (correct) | Seattle Plumbing Co. (correct) |
| Client ID | — | — (correct) | — (correct) | — (correct) | — (correct) | — (correct) | — (correct) | — (correct) | — (correct) | — (correct) | — (correct) |
| Contract type | services | services (correct) | services (correct) | services (correct) | services (correct) | services (correct) | services (correct) | services (correct) | services (correct) | services (correct) | services (correct) |
| Price (net) | 21250.00 | 21250 (correct) | 21250.00 (correct) | 21250.00 (correct) | 21250 (correct) | 21250 (correct) | 21250.00 (correct) | 21250 (correct) | 21250 (correct) | 21250.00 (correct) | 21250.00 (correct) |
| Currency | USD | USD (correct) | USD (correct) | USD (correct) | USD (correct) | USD (correct) | USD (correct) | USD (correct) | USD (correct) | USD (correct) | USD (correct) |
| Billing period | one_time | one_time (correct) | one_time (correct) | one_time (correct) | one_time (correct) | one_time (correct) | one_time (correct) | one_time (correct) | one_time (correct) | one_time (correct) | one_time (correct) |
| Effective date | 2025-05-01 | 2025-05-01 (correct) | 2025-05-01 (correct) | 2025-05-01 (correct) | 2025-05-01 (correct) | 2025-05-01 (correct) | 2025-05-01 (correct) | 2025-05-01 (correct) | 2025-05-01 (correct) | 2025-05-01 (correct) | 2025-05-01 (correct) |
| End date | 2028-12-31 | 2028-12-31 (correct) | 2028-12-31 (correct) | 2028-12-31 (correct) | 2028-12-31 (correct) | 2028-12-31 (correct) | 2028-12-31 (correct) | 2028-12-31 (correct) | 2028-12-31 (correct) | 2028-12-31 (correct) | 2028-12-31 (correct) |
| Notice period (months) | 3 | 3 (correct) | 3 (correct) | 3 (correct) | 3 (correct) | 3 (correct) | 3 (correct) | 3 (correct) | 3 (correct) | 3 (correct) | 3 (correct) |
| Late-payment penalty (% per day) | 0.01 | 0.01 (correct) | 0.01 (correct) | 0.01 (correct) | 0.01 (correct) | 0.01 (correct) | 0.01 (correct) | 0.01 (correct) | 0.01 (correct) | 0.01 (correct) | 0.01 (correct) |
| Governing law | US-NY | US-NY (correct) | US-NY (correct) | US-NY (correct) | US-NY (correct) | US-NY (correct) | US-NY (correct) | US-NY (correct) | US-NY (correct) | US-NY (correct) | US-NY (correct) |
| Tokens | input / output | 877 / 154 | 891 / 151 | 990 / 203 | 1,178 / 387 | 1,291 / 916 | 879 / 179 | 877 / 146 | 879 / 578 | 990 / 141 | 1,242 / 290 |
| of which hidden reasoning | billed as output | 0 | 0 | 95 | 282 | 738 | 0 | 0 | 401 | 33 | 92 |
| of which from cache | input billed at a discount | 403 | 443 | 0 | 654 | 768 | 0 | 0 | 0 | 0 | 0 |
| Cost of this document | this run | $0.0002 | $0.0002 | $0.0002 | $0.0004 | $0.0005 | $0.0007 | $0.0024 | $0.0028 | $0.0034 | $0.0054 |
| Per 1,000 documents | for the same kind of document | $0.17 | $0.19 | $0.20 | $0.38 | $0.52 | $0.71 | $2.41 | $2.83 | $3.39 | $5.38 |
| Latency | this run | 1.2 s | 6.2 s | 2.6 s | 4.6 s | 16.2 s | 1.1 s | 1.2 s | 2.7 s | 2.3 s | 3 s |