Receipt Form Extractor

Status: Live.

1. Your muse

   Rebuild ofFine-tuning Llama models for structured image data extraction (opens in a new tab)Reframe

   Make a tool that pulls structured fields out of photos of receipts and forms, with no fine-tuning. This rebuilds Meta's video "Fine-tuning Llama models for structured image data extraction" (https://dev.meta.ai/llama/resources/videos/llama-fine-tune-structured-image-extraction) on current Muse models. It demonstrates: Customizing Llama to pull structured data out of images. Plan: Same job without fine-tuning. Spark 1.3 reads the image and returns JSON that matches a schema. Upload a receipt or a form and get validated fields back.

2. Show thinking

3. Checks passed

4. 

   The plan: Receipt Form Extractor

   Upload a receipt or form photo and get validated fields as clean JSON.

   1. Upload a photo of a receipt or paper form
   2. Pick a schema such as store receipt or expense form
   3. Run extraction to get structured fields
   4. Review warnings, edit values and copy or download JSON

   See the full plan

   Who it's for

   Freelancers and office staff who digitize receipts and paper forms.

   How it works

   1. Upload a photo of a receipt or paper form
   2. Pick a schema such as store receipt or expense form
   3. Run extraction to get structured fields
   4. Review warnings, edit values and copy or download JSON

   Screens

   • Upload and schema. Upload photo, pick receipt or form type and start extraction.
   • Results and validation. Show fields table and JSON, flag missing values and allow edits and export.

   What goes in

   • Image: One receipt or form photo, JPG or PNG under 4.5 MB. Clear top down shot works best.
   • Text (optional): Optional schema choice such as receipt or expense form plus field list. Defaults provided if empty.

   What comes out

   • JSON: Validated fields such as merchant, date, total, tax and line items plus confidence and warnings.
   • Text: Simple table view of fields for quick review, copy and CSV download.

   Models

   Model	Used for	Why
   Muse Spark 1.3	Read receipt or form photo and return JSON matching schema	Vision model reads the image directly and returns schema matched JSON with no fine tuning.

   Limits

   • Rebuild fit: reframe. Original video fine tuned Llama for extraction, this tool uses Spark 1.3 with preset schemas and no training.
   • Accuracy depends on photo quality. Blur, glare, folds or faint print cause missed fields.
   • Handwriting and unusual layouts are less reliable than printed receipts and standard forms.
   • No connection to accounting, email or storage services. Copy or download JSON manually.

   From your inspo

   Rebuild of Meta video Fine-tuning Llama models for structured image data extraction. Keeps the job of pulling structured data from receipt and form images, now with no fine tuning using Spark 1.3 and schemas.

   Takes

   • Pulling structured data fields out of receipt and form images

   Leaves out

   • Video footage
   • Video script
   • Meta and Llama names and logos

   Cost per run: about $0.005 (est.). About 2000 input tokens at $1.25 per M is $0.0025 plus about 600 output tokens at $4.25 per M is $0.0025 for a total of about $0.005.

6. Done. npm run verify exits 0: typecheck, build, start, console, smoke (6 passed), and manifest all ok.

   It's live.

   musemuse-receipt-form-extractor.vercel.app (opens in a new tab)

   Product pageNew muse

   5 m 36 s · Input 2.2M · Output 49.4k · $0.81 est.