ReductoAI Announces R-1 Document Parsing Model
Team posts emphasize gains in accuracy, speed, and pricing for complex visual documents.
ReductoAI team members posted about launching r-1. Adit described the model as more accurate than the company's strongest agentic OCR versions, quicker to run, and as much as 6x less expensive. The two-year effort focused on handling multi-page tables, varied layouts, and details such as strikethrough text. Raunak added that the tool surpasses LVM parsers for accuracy while costing less than standard OCR options including Azure Doc Intelligence. The announcement was shared by Y Combinator.
Today we’re announcing r-1, our new document parsing model. It’s more accurate than our most powerful agentic OCR models, faster, and up to 6x cheaper. At @reductoai, we spent two years building specialized models for complex visual layouts, tables spanning multiple pages, and…
Reactions
2 postsAD 
Adit@aditabrmToday we’re announcing r-1, our new document parsing model. It’s more accurate than our most powerful agentic OCR models, faster, and up to 6x cheaper. At @reductoai, we spent two years building specialized models for complex visual layouts, tables spanning multiple pages, and…
- likes: 421
- replies: 84
- bookmarks: 218
- reposts: 57
RA Raunak@raunakdoesdevtry this model it's more accurate than LVM parsers and cheaper than basic OCR (e.g. Azure Doc Intelligence), even at our *list price* our ml team absolutely cooked with this one
- likes: 50
- replies: 4
- bookmarks: 16
- reposts: 4