Reducto unveils r-1: a single-pass document parser with 20% error reduction and 1 cent per Page pricing

Reducto has launched r-1, a new single-pass document parsing model that reduces errors by 20% and costs 1 cent per page. The model replaces the company’s previous multi-stage agentic OCR system with a streamlined approach that processes entire pages in one pass. r-1 is available in preview through Reducto’s hosted Parse API on V3, with no open weights or local checkpoints for self-hosting. The model handles digital text, scans, handwriting, and complex layouts, including tables with merged cells and nested headers. Reducto claims r-1 outperforms legacy systems and major hyperscaler products like Amazon Textract and Azure Document Intelligence in internal evaluations. The error reduction is measured against Reducto’s own legacy pipelines, not third-party benchmarks. Pricing is set at 1 cent per page, with no additional costs for high accuracy. The company also announced plans for r-1 mini and automatic per-page routing in the future, with incentives for organizations switching from other parsers.