Tesseract OCR
Ranked 2 of 8 document parsers by recommenders
Free open-source OCR engine, fast and accurate on clean text across many languages. Critics say it struggles with complex layouts and handwriting.
Free open-source OCR engine, fast and accurate on clean text across many languages. Critics say it struggles with complex layouts and handwriting.
60 people gave an opinion
- recommend it
- 32
- criticize it
- 28
- approval share
- 53%
Recommended in 12 of 13 months
- November 2025: recommended
- December 2025: recommended
- January 2026: recommended
- February 2026: recommended
- March 2026: recommended
- April 2026: recommended
- May 2026: recommended
- June 2026: recommended
- July 2026: recommended
- August 2026: recommended
- September 2026: recommended
- October 2026: recommended
- November 2026: not recommended (month in progress)
Why it doesn’t carry the badge
- 32 people recommend itNeeds 10 or more, met.
- 1.1 recommenders for every criticNeeds 2 to 1 or better, not met.
- Recommended in 12 of 13 monthsNeeds 4 or more, met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Fast and runs on CPU
“It's not as accurate as most modern OCRs, but it beats everything else in speed.”
Comment from 14 May 2026 in “The Emacsification of Software” “tesseract is still likely the leading open source OCR that works with CPU”
Comment from 25 Nov 2025 in “Show HN: OCR Arena – A playground for OCR models” “It's not as accurate as I wish, but it's very fast”
Comment from 14 Sep 2026 on Hacker News - Comment from 25 Aug 2026 in “OCR It – pull text out of un-copyable documents for your LLM”
- Comment from 26 Jun 2026 in “LaTeX.wasm: LaTeX Engines in Browsers”
- Comment from 3 Apr 2026 in “Google releases Gemma 4 open models”
Reliable for readable text
“it is the most reliable way to get readable text out”
Comment from 25 Nov 2025 in “Show HN: Stun LLMs with thousands of invisible Unicode characters” “It ended up working quite well.”
Comment from 12 Feb 2026 in “GLM-OCR: Accurate × Fast × Comprehensive” “It works great most of the time”
Comment from 11 Nov 2025 in “I hate screenshots of text” - Comment from 8 Nov 2025 in “Why is Zig so cool?”
Handles multiple languages and diacritics
“It handles non-English characters and diacritics without issues.”
Comment from 25 Aug 2026 in “OCR It – pull text out of un-copyable documents for your LLM” “good for multiple languages, but not very good at handwriting”
Comment from 3 Dec 2025 in “Show HN: A simple Markdown note app with templates and PDF support” Free and open source
“tesseract is still likely the leading open source OCR that works with CPU”
Comment from 25 Nov 2025 in “Show HN: OCR Arena – A playground for OCR models” “Fast free and can run on CPU.”
Comment from 3 Apr 2026 in “Google releases Gemma 4 open models”
What people warn about
Each mark is one comment
Poor layout and column handling
“I thought Tesseract struggled with any documents that weren't composed primarily of a single column of text?”
Comment from 11 Feb 2026 in “GLM-OCR: Accurate × Fast × Comprehensive” “Tesseract does not understand layout. It's fine for character recognition”
Comment from 11 Feb 2026 in “GLM-OCR: Accurate × Fast × Comprehensive” “Tesseract is not a substitute for these models, which understand complex layouts and also extract bounding boxes for things like tables and pictures.”
Comment from 13 Aug 2026 in “Mistral OCR 4.1” Less accurate than newer OCR models
“attention-based OCRs (and more recently VLMs) outperformed any LSTM-based approach since at least 2020”
Comment from 22 Mar 2026 in “25 Years of Eggs” “It's not as accurate as most modern OCRs, but it beats everything else in speed.”
Comment from 14 May 2026 in “The Emacsification of Software” “can 'see' through the noise better than traditional Tesseract-based approaches”
Comment from 2 Mar 2026 in “Seeking Advice on Improving OCR for Watermarked PDFs in My RAG Pipeline” - Comment from 14 Aug 2026 in “Mistral OCR 4.1”
Fragile on faded or noisy scans
“Just noticeably worse character recognition, particularly where the document is faded”
Comment from 21 Aug 2026 in “Anti-AI fonts are useless and harmful” “Sometimes it would do great but small variations could make it fail completely”
Comment from 4 Oct 2026 in “Show HN: AI search for every photo and every frame of video on macOS” “Tesseract is super picky though often failing pixel-perfect screenshots”
Comment from 13 Aug 2026 in “Mistral OCR 4.1” - Comment from 2 Mar 2026 in “Seeking Advice on Improving OCR for Watermarked PDFs in My RAG Pipeline”
Errors need manual correction
“Tesseract's output had a bunch of errors and misdetections, but the main figures always came out OK”
Comment from 13 Aug 2026 in “Mistral OCR 4.1” “I have spent many hours in the current decade correcting OCR errors.”
Comment from 19 Apr 2026 in “Amazon is discontinuing Kindle for PC on June 30th”
Other document parsers people recommend
The four most recommended other document parsers, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| Pandoc | 142 | 28 | 84% | 13 | 10 yrs6 owners | Approved |
| Tesseract OCRThis page | 32 | 28 | 53% | 12 | — | — |
| Docling | 19 | 4 | 83% | 10 | — | Approved |
| Mozilla Readability | 10 | 1 | 91% | 6 | — | Approved |
| Zpdf | 8 | 6 | 57% | 1 | — | — |
What people said
40 of the 64 comments that mention it, newest first. Each links to the original comment.
“now certified fresh without host round trips”
“Sometimes it would do great but small variations could make it fail completely”
“It's not as accurate as I wish, but it's very fast”
“It handles non-English characters and diacritics without issues.”
“Tesseract did better on both accuracy and speed”
“Just noticeably worse character recognition, particularly where the document is faded”
“Tasserect is the worst compare to PaddleOCR , Surya/Marker”
“Tesseract is super picky though often failing pixel-perfect screenshots”
“Tesseract is not a substitute for these models, which understand complex layouts and also extract bounding boxes for things like tables and pictures.”
“Tesseract's output had a bunch of errors and misdetections, but the main figures always came out OK”
“you already get shit ton of accuracy”
“I used to use a custom Tesseract pipeline to OCR the pages”
“select any area of my screen and the text gets OCRed and pops in a buffer”
“I personally just use pipes and tesseract, no need for another tool imo”
“Tesseract is incredibly fast and accurate.”
“No rigorous eval, and I love Tesseract.”
“Most of them had terrible transcriptions on archive.org, made using old tesseract versions.”
“It's not as accurate as most modern OCRs, but it beats everything else in speed.”
“I have spent many hours in the current decade correcting OCR errors.”
“Fast free and can run on CPU.”
“unlike Tesseract which works fine on the scanning hardware I set up for it”
“attention-based OCRs (and more recently VLMs) outperformed any LSTM-based approach since at least 2020”
“can 'see' through the noise better than traditional Tesseract-based approaches”
“And it is completely local using Tesseract”
“Tesseract is so good. Like really good.”
“It ended up working quite well.”
“Tesseract v4 when it was released was exceptionally good and blew everything out of the water.”
“Tesseract does not understand layout. It's fine for character recognition”
“I thought Tesseract struggled with any documents that weren't composed primarily of a single column of text?”
“Teseract supports being trained for specific fonts, that would probably be a good starting point”
“perform OCR (Tesseract and Google Vision)”
“I was surprised that Tesseract had recognized poorly scanned magazines”
“the latter gave me the most granularity when I needed it”
“What makes this practical is that each component runs in its own isolated environment”
“good for multiple languages, but not very good at handwriting”
“tesseract is still likely the leading open source OCR that works with CPU”
“it is the most reliable way to get readable text out”
“also even less prone to those malbehaviors”
“It works great most of the time”
“tesseract does well for me...”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.