Amazon Textract
Shares rank 7 of 11 OCR models by recommenders
Recommended for accurate handwriting recognition and a well-documented Python library; critics call it expensive, prone to surprising costs, and often outperformed by LLMs.
Recommended for accurate handwriting recognition and a well-documented Python library; critics call it expensive, prone to surprising costs, and often outperformed by LLMs.
12 people gave an opinion
- recommend it
- 6
- criticize it
- 6
- approval share
- 50%
Recommended in 5 of 13 months
- November 2025: not recommended
- December 2025: recommended
- January 2026: not recommended
- February 2026: recommended
- March 2026: not recommended
- April 2026: not recommended
- May 2026: recommended
- June 2026: recommended
- July 2026: not recommended
- August 2026: recommended
- September 2026: not recommended
- October 2026: not recommended
- November 2026: not recommended (month in progress)
Why it doesn’t carry the badge
- 6 people recommend itNeeds 10 or more, not met.
- 1 recommender for every criticNeeds 2 to 1 or better, not met.
- Recommended in 5 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
Good handwriting recognition
“it is really good at correctly recognizing hand written characters”
Comment from 5 Dec 2025 in “Gemini 3 Pro: the frontier of vision AI” Well-documented Python library
“great python library with the kind of documentation I expect from a python project”
Comment from 28 Aug 2026 in “I used AWS cognito for a startup. I wouldn't do it again”
What people warn about
Each mark is one comment
Expensive with surprising costs
“They're both very expensive and prone to surprising you.”
Comment from 23 Jun 2026 in “Unlimited OCR: One-Shot Long-Horizon Parsing” “those top of the line traditional OCR systems like AWS, Textract are themselves extremely slow and computationally expensive”
Comment from 26 Dec 2025 in “Rob Pike Goes Nuclear over GenAI” LLMs seen as cheaper alternative
“I don't think I'd do the Textract step since LLMs have gotten way better and cheaper.”
Comment from 3 Nov 2025 in “Tell HN: Mechanical Turk is twenty years old today” “I can replace the amazon textract requests by a llama.cpp model call”
Comment from 25 Apr 2026 in “Show HN: A Karpathy-style LLM wiki your agents maintain (Markdown and Git)” Struggles with dense or odd documents
“Textract did poorly because it was just too dense for their model”
Comment from 14 Aug 2026 in “Mistral OCR 4.1” “a document format that breaks every tool you throw at it”
Comment from 26 Mar 2026 in “We couldn't find an API that understood construction drawings, so we built one”
Other OCR models people recommend
The four most recommended other OCR models, 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 |
|---|---|---|---|---|---|---|
| Mistral OCR | 69 | 40 | 63% | 12 | — | — |
| PaddleOCR | 19 | 4 | 83% | 8 | — | Approved |
| GLM-OCR | 16 | 4 | 80% | 7 | — | Approved |
| DeepSeek-OCR | 12 | 4 | 75% | 9 | — | Approved |
| Amazon TextractThis page | 6 | 6 | 50% | 5 | — | — |
What people said
10 of the 12 comments that mention it, newest first. Each links to the original comment.
“great python library with the kind of documentation I expect from a python project”
“Textract did poorly because it was just too dense for their model”
“tools like these are often worse in the 95 percentile”
“They're both very expensive and prone to surprising you.”
“only if you're using commercial solutions like amazon's textract”
“I can replace the amazon textract requests by a llama.cpp model call”
“a document format that breaks every tool you throw at it”
“those top of the line traditional OCR systems like AWS, Textract are themselves extremely slow and computationally expensive”
“it is really good at correctly recognizing hand written characters”
“I don't think I'd do the Textract step since LLMs have gotten way better and cheaper.”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.