Skip to content

Kingy / Make This

← All outcomes

Extract a checked invoice as JSON

Reproduce one small, fully inspectable extraction task and check every field.

Starting input
The supplied synthetic invoice and public evidence pack.
Result
Reproduce and check every field in one JSON extraction.
Evidence and limits
The accepted output applies to the original input. New invoice variants need their own checks.
Extract a checked invoice as JSON
Retained result from the original project.
Selected model attempts
3
Selected model charges
$0.0000 USD
Human editing
Not recorded

kingy tested

Open the actual result ↗

Original accepted output

{
  "invoice_id": "K-1042",
  "total_usd": 127.50,
  "due_date": "2026-09-20",
  "bank_account": null
}

Qwen3-4B-Q4_K_M.gguf · seed 4101. Retained output for the original input. Changing the recipe does not generate or verify a new result.

Reproduce the outcome

  1. Download and extract the public evidence ZIP.
  2. Run python3 verify_scores.py to recompute the original scores without inference, network requests or downloads.
  3. Inspect raw/local-4B-L01-1.json for the exact request and result.
  4. To make a new invoice variant, copy the adapted prompt and run it only in your chosen local environment.
  5. Compare every field and type with your source invoice; preserve null for absent facts. Record all failures as attempts.
Original input and exact settings
  • Synthetic invoice K-1042, supplied in the exact prompt
temperature
0.6
top_p
0.95
max_tokens
256
context
4096
seeds
[4101,4102,4103]
Original prompt
Return JSON only with invoice_id, total_usd (number), due_date (ISO date), and bank_account (null if absent). Invoice K-1042. Total USD 127.50. Due 2026-09-20. Purchase order PO-88. Do not infer missing facts. /no_think

Attribution and rights: Kingy synthetic task and actual redacted output evidence; model weights remain under their separate license and are not redistributed.

Sources & limits

Small Qwen3 models: retained JSON pilot

Record checked 2026-09-07.

Provider charge was $0. Electricity, hardware and active repair time are unknown.

Only this synthetic invoice was observed. Adapted inputs are untested.

Measured success and timings apply to the exact model/runtime/settings and one M4 Pro.

Help improve this tool

Did this help you finish?

About Make This. Your self-reported outcome goes to Kingy’s private review queue; it is not an independently verified reproduction.

Your experience with Make This
What happened?

All Apps & Labs · Local Lab evidence · Check tool changes

Saved settings stay on this device. Shared links omit private text and actual spending. Reports enter a private review queue. No generated claims or automatic publication.

Contribute a reproducible result

Include the outcome, exact versions, prompts and settings, every attempt, failures, charges if recorded and active repair time. Supply files you own or have permission to share, an attribution and a public evidence link. Remove personal data, tokens and private prompts. Reports are reviewed before any public use; submitting does not publish your work.

Prepare a reproducibility report

Choose an AI setup

Choose your next step

All starting points for this task →