Cohere Parse 5
AssessPlatforms
A multimodal foundation model for extracting structured data from complex documents.
Why it's here
Placed in Assess: 1 article(s) of evidence from 1 source(s), led by product launches, with 0 in the last 30 days. Confidence 24%. Low accumulated evidence, so it defaults conservatively pending more signal.
Evidence (1)
- 7InfoQ·9/3/2026product_launchCohere launches Parse 5 for multimodal document extraction
Cohere has introduced Parse 5, a 2.3-billion-parameter multimodal foundation model for extracting structured information from complex enterprise documents. The model converts visually rich PDFs into Markdown and provides bounding box coordinates for visual grounding, with reported evaluation results on more than 2,000 enterprise pages.