Cohere Releases Parse 5 (parse-v5.0): A 2.3B Vision Language Model That Turns Enterprise Documents Into Markdown

Cohere has launched Parse 5, a 2.3B-parameter vision-language model designed for enterprise document parsing. It converts PDFs, PPTs, and JPEGs into Markdown with text, tables, and layout info, without separate OCR. The model is available via API and cloud platforms, targeting industries with heavy document workflows.