Batch document pipeline
Estimate a high-volume document extraction/summarisation pipeline with batch pricing.
Use this for back-office processing — invoices, statements, claims, contracts — where documents arrive in bulk and a few minutes of latency is acceptable (so batch pricing applies).
Estimate your workload
Model choice is usually the biggest cost lever. Nova Micro/Lite are ~10–70× cheaper than Premier.
Results update automatically as you type.
- Textract OCR(3 pages/doc)
- $450.00
- Bedrock (Amazon Nova Pro)(batch pricing)
- $280.00
- Lambda (triggers)
- $5.02
- DynamoDB (results)
- $1.40
- S3 (documents)
- $2.13
- Per month(all services)
- $738.55
- Per year
- $8,862.60
| Textract OCR | $450.00 | 61% |
| Bedrock (Amazon Nova Pro) | $280.00 | 38% |
| Lambda (triggers) | $5.02 | 1% |
| S3 (documents) | $2.13 | 0% |
| DynamoDB (results) | $1.40 | 0% |
Frequently asked questions
When should I add Textract OCR?+
Only if your documents are scans or images. Digital-text PDFs and DOCX files can be read without OCR, saving the per-page Textract charge.
Why is batch pricing on by default here?+
Bulk document processing rarely needs real-time answers, and Bedrock batch inference is about 50% cheaper. Turn it off if you need synchronous results.
Other AWS architectures
AgentCore and managed-service rates last verified 14 July 2026 (us-east-1). Bedrock model prices last reviewed 28 July 2026. Estimates for planning, not a quote.