Batch Inference

Process a batch of PDFs

Run offline batch processing over a set of PDFs — text extraction, classification, summarization, or structured extraction — in a single job with one receipt.

Run this jobGet help running this

Estimated $0.01–$0.08 per PDF, depending on page count and task · Scales with PDF count and page count

What inputs you need

  • PDF files (or a manifest listing PDF URLs)
  • Processing task (extract/classify/summarize)
  • Model choice

What Badgr returns

  • Per-PDF processed output (text/JSON/summary)
  • Manifest mapping PDF → output
  • Single receipt on completion

Recommended GPU routes

A100 40GBL40S 48GBA40 48GB

Estimated cost: $0.01–$0.08 per PDF, depending on page count and task · Estimated runtime: Scales with PDF count and page count

Example command

badgr run "python batch_pdf.py --input pdfs/ --task extract" \
  --gpu A100_40GB \
  --max-cost 20 \
  --max-runtime 120

Common setup failures Badgr avoids

Scanned/image-only PDFs silently produce empty output — job detects text-layer-less PDFs and routes them through OCR before the requested task.

Job stops partway through a large PDF set — progress is checkpointed per PDF so a restart resumes instead of reprocessing.

Ready to run this?

Launch from the dashboard, CLI, or Compute API. Max-cost protection included.

Run this jobBack to Batch Inference