# Docex

> Agentic vision and OCR infrastructure

## Docs

- [Authentication](https://docs.docex.dev/api-reference/authentication.md): Authenticate API requests with an API key in the x-api-key header. Dashboard access uses GitHub OAuth.
- [Errors](https://docs.docex.dev/api-reference/errors.md): Docex returns structured errors with codes, retryability hints, and actionable suggestions. All errors use the DocexError shape.
- [Runs](https://docs.docex.dev/api-reference/runs.md): Create a vision analysis run, poll for completion, and receive a structured envelope with results, confidence, and usage.
- [Uploads](https://docs.docex.dev/api-reference/uploads.md): Request a presigned upload URL, upload the file directly, then reference the upload ID when creating a run.
- [Changelog](https://docs.docex.dev/changelog/overview.md): Product updates and release notes for Docex.
- [Billing](https://docs.docex.dev/concepts/billing.md): Docex bills at 2× the actual upstream provider cost, estimated before queueing and trued-up after completion. Jobs exceeding $1.00 require confirmation.
- [Pipeline](https://docs.docex.dev/concepts/pipeline.md): The Docex worker pipeline is task-agnostic. It classifies the input, plans a strategy, executes it through AI providers, and returns a structured envelope.
- [Providers](https://docs.docex.dev/concepts/providers.md): Docex routes images and PDFs through an abstraction layer around external AI APIs. The catalog describes capabilities, costs, and auth requirements. The planner resolves candidates. The executor runs strategies with automatic fallback.
- [Schemas](https://docs.docex.dev/concepts/schemas.md): The optional schema registry provides convenience defaults for common extraction tasks. You can bypass it entirely by supplying your own schema at request time.
- [Docex](https://docs.docex.dev/index.md): Agentic vision and OCR infrastructure. Route any image or PDF through the optimal AI pipeline.
- [Agent Playbook](https://docs.docex.dev/integration/agent-playbook.md): Universal setup prompts for coding agents and human developers. Copy-paste these prompts into Codex, Claude Code, Cursor, or OpenCode to scaffold Docex integration.
- [Express](https://docs.docex.dev/integration/express.md): Add a file upload endpoint to an Express app that routes through Docex for vision analysis.
- [Next.js App Router](https://docs.docex.dev/integration/nextjs.md): Scaffold a server-side vision analysis endpoint in a Next.js App Router application using Docex.
- [Introduction](https://docs.docex.dev/introduction.md): Docex is agentic vision and OCR infrastructure. Route any image or PDF through the optimal AI pipeline — with automatic fallback, cost optimization, and structured output.
- [Quickstart](https://docs.docex.dev/quickstart.md): Authenticate with GitHub, top up your wallet, get an API key, and run your first vision analysis — from zero to results in minutes.
- [Installation](https://docs.docex.dev/sdk-cli/installation.md): Install the docexdev SDK and CLI via npm, yarn, or pnpm.
- [Run](https://docs.docex.dev/sdk-cli/run.md): Run vision analysis on an image or PDF using docex.run() or the docex run CLI command.
- [Setup](https://docs.docex.dev/sdk-cli/setup.md): The docex setup command authenticates via GitHub, provisions a wallet, and returns an API key. Designed for agent-first onboarding with human approval at the payment boundary.
- [Verify](https://docs.docex.dev/sdk-cli/verify.md): Accept or correct analysis results for feedback and billing adjustment.
- [Wallet, Jobs & Usage](https://docs.docex.dev/sdk-cli/wallet-jobs-usage.md): Check your wallet balance, list jobs, view usage history, and manage billing from the SDK or CLI.