Cody Baldwin
E-Mail-Text (das Modell hat genau diesen Text erhalten)
From: Cody Baldwin <cody.baldwin@nichols.example> To: support@miller-nguyen.example Date: Wed, Oct 29, 2025 at 4:00 PM Subject: Re: Question about our invoice Dear support team, I'm following up on the message below, nothing has happened so far. Many thanks, we really enjoy working with your team. We need this sorted out by November 23. Kind regards, Cody Baldwin Nichols and Sons On Sun, Oct 26, 2025 at 6:10 PM, support@miller-nguyen.example wrote: > Hello, thank you for your message. We've logged your request as ticket #19305 and are looking into it. > > Miller, Nguyen and Wheeler > On Sun, Oct 26, 2025 at 2:00 PM, Cody Baldwin <cody.baldwin@nichols.example> wrote: > > Dear support team, > > > > We were charged twice for the same payment. Please look into it and refund the overpayment. This concerns invoice #2025-4208. We're talking about 43.87 USD. > > > > Kind regards, > > Cody Baldwin > > Nichols and Sons
Fazit
Am günstigsten ohne Fehler: Gemma 4 31B für 0,13 € pro 1.000 Dokumente.
- Claude Sonnet 5 ist ebenfalls fehlerfrei, aber 30,3× teurer.
- Mistral Small 4 ist 1,4× günstiger, aber: falscher Wert (Priorität).
Prompt für alle Modelle
You triage the support inbox. Read the email below and return only JSON with these keys:
category one of: billing (invoices, payments, charges), delivery (where is the order, shipping,
address change), complaint (damaged, wrong or defective goods, warranty claims),
technical (login, app or system errors, integrations), cancellation (cancel an order,
contract or subscription), sales (quotes, prices, interest before buying)
priority one of: urgent, high, normal, low (rules below)
sentiment one of: positive, neutral, negative (the customer's tone)
customer_name full name of the customer who wrote the request
order_number the customer's order number, if mentioned
invoice_number invoice number, if mentioned
amount the amount of money the request is about (number)
currency ISO 4217 code of that amount (USD, EUR, CZK, …)
deadline the date by which the customer wants a resolution or answer (YYYY-MM-DD)
Priority rules:
- urgent: the customer cannot work right now, or the deadline is at most 1 day after the email date
- high: the deadline is at most 7 days after the email date, or the customer threatens to cancel
or escalate (lawyer, ending the business)
- low: the customer says explicitly that it is not urgent
- normal: everything else
Rules:
- A thread: judge the newest message; order numbers, invoice numbers and amounts may come from the
quoted earlier messages. A forwarded email: the customer is the original sender, not the
colleague who forwarded it.
- Resolve relative deadlines ("by Friday", "by tomorrow", "by the end of the month") against the
date of the message that states them.
- Support ticket numbers, tax IDs and phone numbers are not order or invoice numbers.
- If a value is not in the email, return null. Do not guess.
- Amounts as numbers with a decimal point, no thousands separators and no currency symbol.
So bewerten wir
Exakte Übereinstimmung pro Feld nach Normalisierung von Datumsangaben, Beträgen und Leerzeichen. Fehlende diakritische Zeichen zählen als Fehler. Die Kosten sind der reale Tokenverbrauch dieses Laufs, umgerechnet zum Kurs in der Fußzeile.
Was jedes Modell extrahiert hat
| Feld | Richtiger Wert | Mistral Small 47 / 9 Felder | Gemma 4 31B9 / 9 Felder | GPT-6 Luna8 / 9 Felder | Qwen3.8 Flash9 / 9 Felder | DeepSeek V4.1 Flash9 / 9 Felder | Gemini 3.5 Flash Lite9 / 9 Felder | Mistral Medium 3.57 / 9 Felder | Gemini 3.8 Flash9 / 9 Felder | GPT-6 Sol9 / 9 Felder | Claude Sonnet 59 / 9 Felder |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Kategorie | billing | billing (richtig) | billing (richtig) | billing (richtig) | billing (richtig) | billing (richtig) | billing (richtig) | billing (richtig) | billing (richtig) | billing (richtig) | billing (richtig) |
| Priorität | normal | highfalscher Wert | normal (richtig) | normal (richtig) | normal (richtig) | normal (richtig) | normal (richtig) | highfalscher Wert | normal (richtig) | normal (richtig) | normal (richtig) |
| Stimmung | positive | neutralfalscher Wert | positive (richtig) | neutralfalscher Wert | positive (richtig) | positive (richtig) | positive (richtig) | neutralfalscher Wert | positive (richtig) | positive (richtig) | positive (richtig) |
| Kundenname | Cody Baldwin | Cody Baldwin (richtig) | Cody Baldwin (richtig) | Cody Baldwin (richtig) | Cody Baldwin (richtig) | Cody Baldwin (richtig) | Cody Baldwin (richtig) | Cody Baldwin (richtig) | Cody Baldwin (richtig) | Cody Baldwin (richtig) | Cody Baldwin (richtig) |
| Bestellnummer | — | — (richtig) | — (richtig) | — (richtig) | — (richtig) | — (richtig) | — (richtig) | — (richtig) | — (richtig) | — (richtig) | — (richtig) |
| Rechnungsnummer | 2025-4208 | 2025-4208 (richtig) | 2025-4208 (richtig) | 2025-4208 (richtig) | 2025-4208 (richtig) | 2025-4208 (richtig) | 2025-4208 (richtig) | 2025-4208 (richtig) | 2025-4208 (richtig) | 2025-4208 (richtig) | 2025-4208 (richtig) |
| Betrag | 43.87 | 43.87 (richtig) | 43.87 (richtig) | 43.87 (richtig) | 43.87 (richtig) | 43.87 (richtig) | 43.87 (richtig) | 43.87 (richtig) | 43.87 (richtig) | 43.87 (richtig) | 43.87 (richtig) |
| Währung | USD | USD (richtig) | USD (richtig) | USD (richtig) | USD (richtig) | USD (richtig) | USD (richtig) | USD (richtig) | USD (richtig) | USD (richtig) | USD (richtig) |
| Frist | 2025-11-23 | 2025-11-23 (richtig) | 2025-11-23 (richtig) | 2025-11-23 (richtig) | 2025-11-23 (richtig) | 2025-11-23 (richtig) | 2025-11-23 (richtig) | 2025-11-23 (richtig) | 2025-11-23 (richtig) | 2025-11-23 (richtig) | 2025-11-23 (richtig) |
| Tokens | Eingabe / Ausgabe | 725 / 83 | 765 / 83 | 834 / 164 | 1.080 / 511 | 979 / 454 | 753 / 103 | 725 / 83 | 753 / 471 | 834 / 127 | 1.058 / 214 |
| davon verdecktes Nachdenken | als Ausgabe berechnet | 0 | 0 | 101 | 407 | 396 | 0 | 0 | 368 | 64 | 94 |
| davon aus dem Cache | Eingabe mit Rabatt | 406 | 453 | 0 | 768 | 586 | 0 | 0 | 0 | 0 | 0 |
| Kosten für dieses Dokument | dieser Lauf | 0,0001 € | 0,0001 € | 0,0002 € | 0,0003 € | 0,0004 € | 0,0004 € | 0,0016 € | 0,0021 € | 0,0027 € | 0,0039 € |
| Pro 1.000 Dokumente | für die gleiche Art von Dokument | 0,10 € | 0,13 € | 0,15 € | 0,28 € | 0,36 € | 0,44 € | 1,57 € | 2,14 € | 2,70 € | 3,92 € |
| Latenz | dieser Lauf | 0,7 s | 8,9 s | 2,3 s | 8,4 s | 6,3 s | 0,8 s | 0,9 s | 2 s | 2,9 s | 3,7 s |