Your users upload files — blurry scans, crumpled receipts, low-res screenshots, multi-page PDFs. You need structured data. Docex is the vision layer that handles the models, the OCR, the retries, and the schema validation. One API call. No pipeline to maintain.
Building image-to-data pipelines means evaluating models, handling OCR fallbacks, writing retry logic, and watching costs spiral. Or you lock into one provider and hope it works on every input. Docex is infrastructure — an orchestration engine with an expanding model library that routes each job to the right pipeline, handles failure automatically, and keeps you honest on cost.
Docex maintains a catalog of vision and OCR providers. For each job, it selects the optimal pipeline based on your input, budget, and latency requirements. If a provider fails or returns low confidence, it falls back automatically. You describe what you need in plain text. Docex handles the rest.
await docex.run({ file: "./uploaded-license.heic", prompt: "company name, number, expiry", }); // → 200 OK · 2.4s · ~$0.03 { "legal_name": "ACME LOGISTICS L.L.C", "license_no": "1019388", "expires_on": "2026-03-13" }
Standard document extraction is where most tools stop. Docex starts there and keeps going. Same API. Same structured output. Any input you throw at it.
IDs, licenses, bank statements, proof-of-address
Email attachments, suspicious screenshots, scan reports
Invoices, receipts, purchase orders — crumpled, low-light, any angle
Low-res frame grabs, compressed thumbnails, screen captures
Parties, clauses, signature detection, term extraction
Shipping labels, waybills, manifests, damage photos
Medical records, insurance claims, lab results, handwritten notes
This is what we've tested. Docex adapts to any image-to-data task you have in store.
We charge 2× what the upstream provider charges us. No markup games, no opaque credits. For most inputs, that means pennies per request. Dynamic routing helps — when a cheaper OCR model handles the job, you pay less. Drop in five dollars to start. Cancel anytime. No annual contract, no sales call.
Paste this into your coding agent. It wires Docex into your product, scaffolds the endpoint for your stack, and runs a smoke test. You approve one link. Production-ready vision analysis without the production-ready team.
Wire Docex into this project as the vision analysis layer for [describe the use case — e.g., "reading trade licenses during user onboarding"]. Take me through the GitHub approval and the $5 wallet top-up, store the API key in my env, scaffold a server-side analysis endpoint for my stack, and run a smoke test to confirm the integration works end-to-end.