Benchmark Almanac

OCR Arena

Gemini 3 Flash 1,781 leads since December 2025
Not in model cards

Users upload a document, compare two anonymous OCR or vision-language models side by side, and vote. Votes become Elo ratings.

Launched with more than 10 models on Hacker News on November 21st, 2025 (216 points).

Made by
Extend
Released
November 2025
Kind
Documents, charts and text in images
Level
Everyday
Scored
Ranked by votes
Answers
Open-ended
Tasks
All public
Strongest model tested
Claude Opus 4.6 #37 on the capability index
Last model added
February 2026
Scores refreshed
November 12th, 2026
18
Entries on its own leaderboard, 11 models we track

Gemini 3 Flash leads with 1,781 among models released by February 2026

Best score per entry from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.

The newest model with a score here came out in February 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

18 entries. Scroll sideways for the rest. Bars start at 1,300.

  1. 1,781 Gemini 3 Flash
  2. 1,778 Claude Opus 4.5
  3. 1,745 Gemini 2.5 Pro
  4. 1,709 Gemini 2.5 Flash
  5. 1,679 Claude Opus 4.6
  6. 1,631 Claude Sonnet 4.5
  7. 1,530 Qwen3-VL-8B
  8. 1,515 dots.ocr
  9. 1,512 Mistral OCR v3
  10. 1,486 Nanonets2-3B
  11. 1,438 GLM-OCR
  12. 1,422 olmOCR 2
  13. 1,406 GPT-5
  14. 1,377 DeepSeek OCR
  15. 1,367 GPT-5.1
  16. 1,367 Gemini 3 Pro
  17. 1,356 GPT-5.2
  18. 1,319 Iris

Bars start at 1,300.

  1. 1 Gemini 3 Flash 1,781
  2. 2 Claude Opus 4.5 low 1,778
  3. 3 Gemini 2.5 Pro Jun 2025 1,745
  4. 4 Gemini 2.5 Flash Sep 2025 1,709
  5. 5 Claude Opus 4.6 adaptive 1,679
  6. 6 Claude Sonnet 4.5 1,631
  7. 7 Qwen3-VL-8B 1,530
  8. 8 dots.ocr 1,515
  9. 9 Mistral OCR v3 1,512
  10. 10 Nanonets2-3B 1,486
  11. 11 GLM-OCR 1,438
  12. 12 olmOCR 2 1,422
  13. 13 GPT-5 low 1,406
  14. 14 DeepSeek OCR 1,377
  15. 15 GPT-5.1 medium 1,367
  16. 16 Gemini 3 Pro 1,367
  17. 17 GPT-5.2 none 1,356
  18. 18 Iris 1,319

Source: Official OCR Arena leaderboard

#EntryScoreSettingDateSource
1 Gemini 3 Flash 1,781 Dec 2025 Official leaderboard
2 Claude Opus 4.5 1,778 low Nov 2025 Official leaderboard
3 Gemini 2.5 Pro Jun 2025 1,745 Jun 2025 Official leaderboard
4 Gemini 2.5 Flash Sep 2025 1,709 Sep 2025 Official leaderboard
5 Claude Opus 4.6 1,679 adaptive Feb 2026 Official leaderboard
6 Claude Sonnet 4.5 1,631 Sep 2025 Official leaderboard
7 Qwen3-VL-8B 1,530 Official leaderboard
8 dots.ocr 1,515 Official leaderboard
9 Mistral OCR v3 1,512 Official leaderboard
10 Nanonets2-3B 1,486 Official leaderboard
11 GLM-OCR 1,438 Jan 2026 Official leaderboard
12 olmOCR 2 1,422 Official leaderboard
13 GPT-5 1,406 low Aug 2025 Official leaderboard
14 DeepSeek OCR 1,377 Official leaderboard
15 GPT-5.1 1,367 medium Nov 2025 Official leaderboard
16 Gemini 3 Pro 1,367 Nov 2025 Official leaderboard
17 GPT-5.2 1,356 none Dec 2025 Official leaderboard
18 Iris 1,319 Official leaderboard

Scores at each effort level

Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.

Lowest effort to highest, light to dark
Claude Opus 4.5
low medium
1,778
Claude Opus 4.6
adaptive
1,679
GPT-5
low medium
1,406
GPT-5.1
low
1,367
GPT-5.2
none
1,356

Source: Official OCR Arena leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Gemini 3 Flash 1,781
Claude Opus 4.5 low 1,778
Gemini 2.5 Pro Jun 2025 1,745
Gemini 2.5 Flash Sep 2025 1,709
Claude Opus 4.6 adaptive 1,679
Claude Opus 4.6 medium 1,666
Claude Opus 4.6 low 1,660
Claude Opus 4.5 medium 1,658
Claude Sonnet 4.5 1,631
GLM-OCR 1,438
GPT-5 low 1,406
GPT-5.1 medium 1,367
Gemini 3 Pro 1,367
GPT-5 medium 1,366
GPT-5.1 low 1,366
GPT-5.2 none 1,356
GPT-5.2 medium 1,330

3 models have held number one on the official leaderboard since its release in November 2025

Lab claims count from the day the card came out, independent runs from the model’s release.

Google DeepMind Anthropic
Own leaderboard Not tested since Feb 2026
Gemini 3 Flash 1,781

Source: Official OCR Arena leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

On the benchmark’s own leaderboard
1,300 1,400 1,500 1,600 1,700 1,800 2025 2026 1,300 1,400 1,500 1,600 1,700 1,800 2025 2026

Source: Official OCR Arena leaderboard

No lab model card reports it yet

We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.

More in Vision

All 281 in Vision
  • MMMU In use 11.5K college-level multimodal questions across 30 subjects. Multiple-choice and open accuracy. GPT-5.1 85.4%
  • MMMU-Pro In use Harder MMMU: college-level multimodal questions with more options and vision-only inputs to remove text shortcuts. GPT-5.6 Sol 84.6%
  • CharXiv In use Chart understanding on real arXiv figures. Reasoning questions require synthesizing information from complex charts. Claude Fable 5 93.5%
  • MATH-Vision (MathVision) In use 3,040 competition math problems with visual context. Answer accuracy. Qwen 3.8 Max 97.7%
  • MMStar In use Models answer 1,500 human-curated image questions that cannot be solved without the image, across 6 core capabilities. Scored by accuracy. Qwen 3.8 Max 85.9%
  • ChartQA In use Models answer human-written and generated questions about chart images that need visual and arithmetic reasoning. Scored by relaxed accuracy with a 5 percent numeric tolerance. Command A Vision 90.9%