Field noteOct 23, 20261 min read1 source

OCRia-1 document parsing model from LlamaIndex is now available on OpenDocRouter

In one minute

OpenDocRouter’s models catalog now lists LlamaIndex’s document-parsing model as `llamaindex/ocria-1`, versioned 2026-10-22.

What changed: OCRia-1 is now available through OpenDocRouter, with listed benchmark scores, page cost and speed.

The service says models are selected by passing their ID in a parse request and that each version has a parsing recipe containing its prompt and instructions.

Its ParseBench table reports an overall score of 84.67 for OCRia-1, with 91.24 for tables and 80.95 for charts. The page estimates $13.04 per 1,000 typical pages and a typical processing time of 2.0 seconds per page. Listed token prices are $0.97 per million input tokens, $0.97 per million cached input tokens, and $4.83 per million output tokens; the maximum charge per page is $0.180.

OpenDocRouter permits up to 50 pages per synchronous request and 500 per asynchronous request. The page says benchmark cost estimates use ParseBench documents and may differ for a customer’s pages. It does not establish OCRia-1’s performance, cost, or latency on any particular document set, nor does it provide an independent evaluation of the model.

Evidence / sources

  1. 1opendocrouter.ai/models
All Field Notes

Next reads

Read next

Open the graph
  1. 01topicModelsExplore the models topic hub.
  2. 02archiveField NotesOpen the latest editorial analysis.
  3. 03source ledgerSource LedgerInspect the public source evidence graph.
Who read this page?Machine requests, hidden until opened

Loading the privacy-safe route aggregate...

Open the JSON contract