Meta tag

Omnidocbench

2 articles

Baidu’s 4B OCR marries vision and language

AIRewritten

db#2881

Baidu’s model reads the whole document, not just the text on the scan

Baidu's Qianfan team released a 4-billion-parameter model that collapses layout analysis, text recognition, and document understanding into a single end-to-end neural stack.

18 Mar 2026

$FireRed-OCR-2B tries to stop AI from inventing tables$

AIRewritten

db#3809

Document AI is moving from reading text to preserving the structure around it

FireRedTeam’s FireRed-OCR-2B model scores 92.94% on OmniDocBench v1.5, outperforming rivals by enforcing LaTeX and table syntax rules during training.

02 Mar 2026

Meta tag

Omnidocbench

2 articles

AIRewritten

db#2881

Baidu’s model reads the whole document, not just the text on the scan

Baidu's Qianfan team released a 4-billion-parameter model that collapses layout analysis, text recognition, and document understanding into a single end-to-end neural stack.

18 Mar 2026

$FireRed-OCR-2B tries to stop AI from inventing tables$

AIRewritten

db#3809

Document AI is moving from reading text to preserving the structure around it

FireRedTeam’s FireRed-OCR-2B model scores 92.94% on OmniDocBench v1.5, outperforming rivals by enforcing LaTeX and table syntax rules during training.

02 Mar 2026