Intelligent document processing · November 2026 Edition
Google Document AI vs Amazon Textract
Zero of fourteen models named Google Document AI first on the direct prompt; zero named Amazon Textract. Google Document AI was named by thirteen of the fourteen models and Amazon Textract by twelve and Google Document AI carries 29 labels and Amazon Textract 27, so the shares are not directly comparable.
Google Document AI
accepted challengerNamed in one category this edition.
Amazon Textract
criticized challengerNamed in one category this edition.
First-choice share4%4%Of first choices across the direct, paraphrase, budget and scale prompts, 0 to 100.
Negative rate24%30%Negative labels as a share of the product's labels, 0 to 100.
Rank in category#7#8A position in a field of 13; printed, not drawn.
Labels2927A count; the two differ.
The two percentage rows are drawn on one 0 to 100 track, Google Document AI reading right to left. Rank and label count are printed, not drawn.Rossum was named alongside these two in eleven of the fourteen direct answers. Nanonets vs Google Document AI · Nanonets vs Amazon Textract · Rossum vs Google Document AI
Share is the count of first choices across the direct, paraphrase, budget and scale prompts over all fourteen models, for a mid-market B2B company; rank is within the category; every quote names the model and the prompt it came from. Both figures come from the intelligent document processing page.
By framing
How many of the fourteen models made each the first choice, per way of asking, and how many argued against it.
Google Document AIFirst choices, of fourteen modelsAmazon Textract
Direct001 against Google Document AI · 1 against Amazon Textract
Paraphrase00
Comparative001 against Amazon Textract
Budget-constrained221 against Google Document AI · 2 against Amazon Textract
Scale-constrained00
Negative005 against Google Document AI · 4 against Amazon Textract
Bars are first choices, 0 to 14 each sideModels that argued againstA model can name both, so the two sides of a row do not sum to fourteen.
Across every category in the November 2026 Edition, Google Document AI and Amazon Textract were named in the same answer eighty times, of the 92 answers naming Google Document AI and the 89 naming Amazon Textract. In those answers Amazon Textract took the first choice one time and Google Document AI zero.
Every model, every framing
The eighty-four answers behind the chart above, one cell each: where Google Document AI and Amazon Textract stood in it.
| Model | DirectGDAT | ParaphraseGDAT | ComparativeGDAT | Budget-constrainedGDAT | Scale-constrainedGDAT | NegativeGDAT |
|---|---|---|---|---|---|---|
| Claude Haiku 4.5 | GDAT | |||||
| GPT-5.4 mini | GDAT | |||||
| Gemini 3.5 Flash | GDAT | GDAT | GDAT | GDAT | GDAT | |
| Perplexity Sonar | GDAT | AT | GDAT | |||
| Grok 4.20 | GD | |||||
| Mistral Small | GD | |||||
| DeepSeek V4 Flash | GDAT | |||||
| Llama 4 Maverick | GD | |||||
| Qwen 3.7 Flash | GDAT | GDAT | GDAT | |||
| Kimi K2 | GDAT | |||||
| GLM 4.7 FlashX | ||||||
| MiniMax M2.5 | ||||||
| GPT-6 Luna | GDAT | GDAT | ||||
| Muse Glimmer 30B | GD |
GD Google Document AIAT Amazon TextractGD first choiceGD named as an alternativeGD argued againstblank: not namedEach cell is one answer, Google Document AI on the left and Amazon Textract on the right.
The direct prompt
The plain question, one answer per model, grouped by where Google Document AI and Amazon Textract stood in it.
Neither was the first choice, one was named
1 of 14 modelsThe answer put something else first and named one of the two as an alternative.Muse Glimmer 30BNanonets, Rossum alternatives: ABBYY Vantage, Automation Anywhere, Google Document AI, UiPath Document Understanding
Neither was named
13 of 14 modelsThe answer made no first choice from these two in this category.Claude Haiku 4.5Nanonets alternatives: Doxis AI.dp, Rossum
GPT-5.4 miniABBYY Vantage alternatives: AWS IDP, Azure AI Document Intelligence, Rossum
Gemini 3.5 FlashNanonets alternatives: ABBYY Vantage, Docsumo, Rossum
Perplexity SonarNanonets alternatives: Docsumo, Microsoft Azure AI Document Intelligence, Parseur
Grok 4.20Rossum alternatives: ABBYY Vantage or FlexiCapture, Docsumo, Hyperscience Hypercell, UiPath Document Understanding
Mistral SmallABBYY Vantage, Rossum alternatives: Apryse, Turian
DeepSeek V4 FlashDocsumo alternatives: Nanonets, Rossum
Llama 4 MaverickNanonets alternatives: Apryse PDF SDK, Parseur
Qwen 3.7 FlashNanonets alternatives: Extend, Rossum, Turian
Kimi K2Rossum alternatives: ABBYY Vantage, Microsoft Azure AI Document Intelligence, Nanonets
GLM 4.7 FlashXUiPath Document Understanding alternatives: ABBYY Vantage, Automation Anywhere Intelligent Automation, Microsoft Azure AI Document Intelligence, ServiceNow Document Management
MiniMax M2.5Nanonets alternatives: ABBYY Vantage, Docsumo, Hyperscience Hypercell, Rossum, UiPath Document Understanding
GPT-6 LunaRossum alternatives: ABBYY Vantage, Nanonets, UiPath Document Understanding
Bold names in an answer are the products the judge labeled a first choice; a model naming several gives each of them that label. The full answer text for every row is in the record.
By buyer segment
The same question asked on behalf of a different buyer. Each standing is computed within its segment and they are never added together. The figures above are the mid-market standing, which is the one the category orders by.
Small business
Level: the same share of first choices.
Google Document AI0%#7 of 12
Amazon Textract0%#9 of 12
The full small business standing →Mid-marketThe figures above
Level: the same share of first choices.
Google Document AI4%#7 of 13
Amazon Textract4%#8 of 13
The full mid-market standing →Enterprise
Level: the same share of first choices.
Google Document AI0%#11 of 11
Amazon Textract0%#10 of 11
The full enterprise standing →What the models said about Google Document AI
Every negative label with a quote, up to three, then the highest-weighted positives, up to three. Six of eight in this category shown.
“Avoid if you do not have a dedicated DevOps team to maintain the code.” Qwen 3.7 Flash · negative prompt · hard negative
“Google Document AI, Amazon Textract, and Azure Document Intelligence can provide strong extraction, but you may need to build validation, exception queues... They are risky choices” Perplexity Sonar · negative prompt · soft negative
“Basic cloud extraction engines (e.g., basic AWS Textract, Google Document AI base models, Azure AI Document Intelligence).” Gemini 3.5 Flash · negative prompt · soft negative
“"I have a developer" | Amazon Textract or Google Document AI | Rawest price per page ($0.0015). No markup.” Qwen 3.7 Flash · budget prompt · first choice
“such as Azure AI Document Intelligence, Google Document AI, or Nanonets” Llama 4 Maverick · budget prompt · first choice
“For more technical users, Google Document AI offers a pay-as-you-go model with transparent pricing.” Mistral Small · budget prompt · alternative
What the models said about Amazon Textract
Every negative label with a quote, up to three, then the highest-weighted positives, up to three. Six of eight in this category shown.
“Avoid if you do not have a dedicated DevOps team to maintain the code.” Qwen 3.7 Flash · negative prompt · hard negative
“Amazon Textract as a standalone, turnkey IDP platform. Treat it as a document-processing building block: you may need to add workflow orchestration, validation, and exception handling yourself.” GPT-6 Luna · negative prompt · soft negative
“Google Document AI, Amazon Textract, and Azure Document Intelligence can provide strong extraction, but you may need to build validation, exception queues... They are risky choices” Perplexity Sonar · negative prompt · soft negative
“"I have a developer" | Amazon Textract or Google Document AI | Rawest price per page ($0.0015). No markup.” Qwen 3.7 Flash · budget prompt · first choice
“Amazon Textract is often the safest budget-conscious choice for teams that want a scalable IDP foundation” GPT-5.4 mini · budget prompt · first choice
“Developers usually prefer AWS Textract or Azure AI Document Intelligence APIs to integrate into custom apps” Gemini 3.5 Flash · comparative prompt · alternative
Also compared
Nanonets vs Google Document AINanonets vs Amazon TextractRossum vs Google Document AIRossum vs Amazon TextractDocsumo vs Google Document AIDocsumo vs Amazon TextractParseur vs Google Document AIParseur vs Amazon TextractABBYY Vantage vs Google Document AIABBYY Vantage vs Amazon TextractUiPath Document Understanding vs Google Document AIUiPath Document Understanding vs Amazon TextractAll twenty-eight head-to-head pages and the full document processing standing
Comparisons are drawn for the top eight products in each category, each against each. The output is the models' output; nothing here is a recommendation by the index.