Extract selected invoice fields in your own AWS account, offline on CPU. Preserve source decimal strings and explicit missing values. Use XML for structured extraction and supported printed or scanned PDFs for transcription with required review.
Software: USD 0.50 per instance-hour on ml.c5.2xlarge. Eligible buyers receive a 14-day software trial. AWS infrastructure, storage and data-transfer charges are separate, including during the trial. Consult the Marketplace offer for final terms.
Supported preview
Mexico: CFDI 3.3/4.0 I/E XML and bounded CFDI printed PDF layouts.
Brazil: legacy NF-e 4.00 model 55 XML, bounded DANFE PDFs, and a narrow provider-issued national NFS-e 1.01 XML subset with embedded DPS.
English, Spanish and Portuguese local OCR artifacts; no external OCR service or request-time weight downloads.
Every PDF requires comparison with its source. This product transcribes fields; it does not verify fiscal validity, signatures, issuers, taxpayer identity, government acceptance or compliance. Multiline tables, handwriting, arbitrary rotation, municipal NFS-e variants and new tax-reform groups are outside this preview.
The finite synthetic evaluation passes its declared objectives, with retained errors including incorrect identifiers. See the model card, usage, sources and self-contained notebook.
Support: info@softwaresushi.com. Send error codes and redacted structure; keep taxpayer records and banking data out of support messages.
Watch the product workflow
See a supported Mexican invoice XML and a fictional PDF preview processed by the actual published model. The walkthrough includes source comparison; every PDF requires review and extraction does not establish fiscal validity.
Watch the English product video on YouTube. The interface is an illustrative integration rather than an included product UI. The examples are fictional; model responses are real.
Highlights
Exact decimal strings, preserved XML tax paths and per-field PDF provenance
Offline CPU OCR with locally pinned Apache-2.0 language weights
Conservative missing values and required review; no fiscal validation
AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
You pay by the hour based on how long the software runs on an ml.c5.2xlarge instance. The two dimensions split by processing mode. Real-time inference handles requests as they arrive for live extraction. Batch inference processes groups of documents together. Both use the same instance type, so your hourly rate depends only on which mode you select. Costs scale with the number of hours each mode runs. AWS infrastructure and data transfer are billed separately from this software charge.
Top-of-mind questions for buyers
What does the ml.c5.2xlarge instance type mean for real-time and batch inference?
The ml.c5.2xlarge is a CPU-based compute instance that runs the extraction model. Both dimensions use this same instance size. Your hourly software charge reflects how long that instance runs, regardless of whether you chose the real-time or batch mode.
When should I pick real-time inference versus batch inference?
Real-time inference processes each document request as it arrives, suited to live extraction. Batch inference groups documents and processes them together, suited to scheduled runs. Both bill per hour the instance runs. You pick the mode that fits your document flow.
What costs are separate from the software charge?
AWS infrastructure such as EC2, EBS storage, networking, and data transfer is billed separately. The software charge meters only the time the model runs on the ml.c5.2xlarge instance. Review the current listing before subscribing to confirm these separate charges.
softwaresushi.com
Helpful?
Vendor refund policy
Software charges stop when the endpoint is deleted or a batch job ends. If the product did not work as described, email info@softwaresushi.com within 30 days with your AWS account ID and endpoint or job name. Software Sushi will refund the software charges for that use. AWS infrastructure charges are separate and are not covered.
How can we make this page better?
Tell us how we can improve this page, or report an issue with this product.
Give us feedbackReport a problem with this product or seller
Legal
Vendor terms and conditions
Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA).
Content disclaimer
Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.
An Amazon SageMaker model package is a pre-trained machine learning model ready to use without additional training. Use the model package to create a model on Amazon SageMaker for real-time inference or batch processing. Amazon SageMaker is a fully managed platform for building, training, and deploying machine learning models at scale.
Deploy the model on Amazon SageMaker AI using the following options:
Real-time inference
Deploy the model as an API endpoint for your applications. When you send data to the endpoint, SageMaker processes it and returns results by API response. The endpoint runs continuously until you delete it. You're billed for software and SageMaker infrastructure costs while the endpoint runs. AWS Marketplace models don't support Amazon SageMaker Asynchronous Inference. For more information, see Deploy models for real-time inference .
Batch transform
Deploy the model to process batches of data stored in Amazon Simple Storage Service (Amazon S3). SageMaker runs the job, processes your data, and returns results to Amazon S3. When complete, SageMaker stops the model. You're billed for software and SageMaker infrastructure costs only during the batch job. Duration depends on your model, instance type, and dataset size. AWS Marketplace models don't support Amazon SageMaker Asynchronous Inference. For more information, see Batch transform for inference with Amazon SageMaker AI .
Version release notes
Maintained Ubuntu26.04 runtime with vendor security updates, unchanged pinned recognition weights and all33200 synthetic target fields preserved. Passes original XML/PDF objectives and320 container tests. Exact-image realtime AWS serial84/84 succeeds. Every PDF requires review; finite synthetic results do not establish fiscal validity or broad customer accuracy. Narrow national NFS-e1.01 retains its disclosed reference-XSD series limitation.
Additional details
Inputs
Outputs
Usage instructions
Sample notebooks
Inputs
Summary
JSON object with base64-encoded XML or PDF; JSON Lines; raw application/xml or application/pdf.
Limitations for input type
1 MiB XML; 4 MiB PDF; 6 MiB HTTP body; 16 records and at most 2 PDFs per request; 3 PDF pages. Only provider-issued national NFS-e 1.01 XML subset supported. See usage for resource bounds.
Contact info@softwaresushi.com with redacted structure and error codes only. Never send taxpayer data in public issues.
AWS infrastructure support
AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.
pdfRest API Toolkit delivers powerful, secure, and scalable PDF processing, developed by the document experts at Datalogics.
With dozens of flexible API endpoints, it supports document conversion, editing, security controls, data extraction, redaction, XFA forms and Acroform handling, PDF to Office exports, and more. Easily integrate into any application or workflow to automate bulk, batch, and real-time document processing with full control via a dedicated EC2-based REST API.
Built on the Adobe® PDF Library™ - the same core technology that powers Adobe Acrobat - pdfRest offers the gold-standard solution for high-quality, customizable PDF automation trusted by developers and enterprises worldwide.
Run the full pdfRest PDF processing API as a self-hosted container in your own AWS account - 40+ REST API endpoints, entirely inside your VPC. Bring your own pdfRest license.
Be the first to review this product. We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.