built by extend.ai

GPT-5.1 (Medium) vs Nanonets2-3B

Head-to-head OCR comparison · 91 battles · Updated May 23, 2026

GPT-5.1 (Medium)
Nanonets2-3B
Matchup72.5%(66W-25L)27.5%(25W-66L)
Overall52.8%(1216W-832L)29.0%(450W-905L)
ELO15041415
Rank#12#19

In head-to-head OCR battles, GPT-5.1 (Medium) outperforms Nanonets2-3B with a 72.5% win rate across 91 matchups. GPT-5.1 (Medium) currently ranks #12 on the OCR Arena leaderboard with an ELO rating of 1504, while Nanonets2-3B holds position #19 with an ELO of 1415. These rankings are based on blind community voting where users compare OCR outputs without knowing which model produced each result.

Try it yourself

Test these models on your own documents in the playground.

Open Playground →

Other comparisons

GPT-5.1 (Medium) vs Nanonets2-3B - OCR Model Comparison | OCR Arena