← Back to People
Open-sourced Chandra OCR 2, achieving 85.9% on olmOCR bench with 90+ language support and a smaller 4B model.
4
Mentions
3
Articles
2
Outlets
Mar 2026 — Oct 2026
Coverage
Positive
Sentiment
Companies Covered
Appears In
Sentiment Distribution
Based on 3 scored articles
Positive 100%
Articles
Most recent firstArticles Written
Vik Paruchuri as author
Also Mentioned In
Referenced in coverage
Indra OCR 1.1 Open-Sourced by Datalab
DatalabpositiveOct 25, 2026
“Author of the Indra OCR 1.1 release article describing an open-source 1M-parameter OCR model released by Datalab.”
Surya OCR 2
SubstackpositiveMay 28, 2026
“Announced Surya OCR 2, a 650M parameter model achieving 83.3% on OLMOCR benchmark and 87% on internal 91-language benchmark.”