Parse documents and transcribe audio in one API.

Send PDFs, scans, office files or call recordings and get back one structured format. Every block carries page coordinates or audio timecodes, and every extracted field points to its source.

Four primitives to read any file.

Upload a file once and reuse it across Parse, Split, Extract and Transcribe so you never pay to parse the same document twice.

Parse once, extract from documents and calls.

Parse a scanned packet and a call recording, then pass both into Extract under a single schema.

Connect your coding agent

Works in Claude Code, Codex and Cursor

  1. Reads the CloudRaker docs and installs the Paperwork skills.
  2. Maps where documents, calls and signatures move through your code and process.
  3. Returns a ranked integration plan and asks before it writes code.
Preview the prompt
Help me integrate CloudRaker, the Paperwork platform, into this project.
Read https://docs.cloudraker.com/llms.txt. Append .md to any docs page for Markdown.
Install the Paperwork skills for the coding agent you are running in:
npx skills add cloudraker/cloudraker-paperwork-skills --agent <client> --skill '*' --yes
Choose claude-code, codex, or cursor for <client>. The MCP server at mcp.cloudraker.com serves the whole API.

1. Read this codebase and the process it supports. Find where documents, scans, call recordings, forms or signatures enter or leave it.
2. For each place, name the primitive that fits (parse, split, classify, extract, transcribe, decide, fill, sign) and the surface: the Python SDK, the TypeScript SDK, the paperwork CLI, or MCP. Say which file or step changes and what it replaces.
3. Flag manual steps that one call could replace, even where no code exists yet.
4. Return a ranked list with location, primitive, rationale and effort. Ask me where to start before you write code.
5. API keys are self-serve at https://signup.cloudraker.com, with $15 of free usage a month. Never print keys, and keep them out of the repository and chat.

Context: Focus on the Understand stage: Parse documents and call recordings once, then Extract under one schema.

80.1 on ParseBench, 270x realtime on audio.

rakedoc-nano-2 ranks first among open-weight parsers on ParseBench and fifth overall (LlamaParse Agentic Plus leads overall at 86.7). rakeaudio-asr transcribes a 10-minute call in 2.2 seconds at ~5% internal English WER (Gemini 3.5 Transcribe leads on accuracy at 2.6%; Deepgram Nova-3 is faster at 300x).

Understand rates and limits.

1 credit is $0.001. Every account starts with $15 of free usage a month and no card.

RateScope and limits
Parse (Markdown)1 credit ($0.001) / pageBorn-digital PDFs, markdown output only
Parse (Rule-based JSON)2 credits ($0.002) / pageStructured JSON and bounding boxes, no GPU needed
Parse (OCR / rakedoc-nano-2)10 credits ($0.01) / page ($5.20/1,000 Enterprise)Scanned or dense pages; replaces base parse fee
Split2 credits ($0.002) / page; 4 credits ($0.004) / audio minUp to 750 pages; dry runs are free
Extract15 credits ($0.015) / unit Smart; 30 credits ($0.03) AdvancedUp to 100 files per callAdd 5 credits/unit for citations and 5 credits/unit for scoring.
Transcribe5 credits ($0.005) / min; 20 credits ($0.02) / min diarized24 languages; batch and live WebSocket streamingRealtime streaming: 5 credits/session + 2 credits/started min.

Next in the chain: Decide.

Once a file is parsed or transcribed, send it to Decide to classify packets or branch your code on a measured probability in milliseconds. Under the hood, Understand runs on two CloudRaker models: rakedoc-nano-2 (1.2B open weights, AGPL-3.0) for scanned pages and rakeaudio-asr (270x realtime) for speech.

Send your hardest PDF or call recording.

$15 of free usage a month covers 1,500 scanned pages or 3,000 minutes of transcription, with no card required.