Zero of fourteen models named Google Document AI first on the direct prompt; zero named Amazon Textract. Google Document AI was named by twelve of the fourteen models and Amazon Textract by thirteen and Google Document AI carries 23 labels and Amazon Textract 28, so the shares are not directly comparable.
Named in one category this edition.
Named in one category this edition.
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.
Across every category in the October 2026 Edition, Google Document AI and Amazon Textract were named in the same answer fifty-seven times, of the 78 answers naming Google Document AI and the 77 naming Amazon Textract. In those answers Amazon Textract took the first choice zero times and Google Document AI two.
| Model | DirectGDAT | ParaphraseGDAT | ComparativeGDAT | Budget-constrainedGDAT | Scale-constrainedGDAT | NegativeGDAT |
|---|---|---|---|---|---|---|
| Claude Haiku 4.5 | AT | GD | AT | |||
| GPT-5.4 mini | AT | GD | ||||
| Gemini 3.5 Flash | GDAT | GDAT | ||||
| Perplexity Sonar | AT | AT | ||||
| Grok 4.1 Fast | GDAT | GDAT | ||||
| Mistral Small | ||||||
| DeepSeek V4 Flash | GDAT | GDAT | GD | |||
| Llama 4 Maverick | ||||||
| Qwen 3.7 Flash | GDAT | AT | ||||
| Kimi K2 | ||||||
| GLM 4.7 FlashX | GDAT | |||||
| MiniMax M2.5 | AT | GDAT | ||||
| GPT-6 Luna | GD | GDAT | GD | AT | ||
| Muse Glimmer 30B | GDAT | AT |
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.
Every negative label with a quote, up to three, then the highest-weighted positives, up to three. Six of seven in this category shown.
“center on managed analysis APIs, so buyers assemble most of the operational queue, review, routing, and export application around them” Claude Haiku 4.5 · comparative prompt · soft negative
“Users report pricing issues that hinder adoption for smaller businesses; costs can escalate unpredictably” DeepSeek V4 Flash · negative prompt · soft negative
“Google Cloud Document AI” Gemini 3.5 Flash · negative prompt · soft negative
“My practical pick: Google Document AI if you want a strong cloud-native option with a free tier and transparent usage pricing.” GPT-5.4 mini · budget prompt · first choice
“Go with Google Cloud Document AI or Azure AI Document Intelligence. Their free tiers are highly generous” Gemini 3.5 Flash · budget prompt · first choice
“I would suggest starting with Google Cloud Document AI or Amazon Textract” MiniMax M2.5 · budget prompt · first choice
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 are non-technical: Azure AI and AWS Textract” Qwen 3.7 Flash · negative prompt · hard negative
“Be especially cautious if... custom extraction is developer-centric and that handwriting and language support need testing.” GPT-6 Luna · negative prompt · soft negative
“"Raw" Cloud Provider APIs (Without a Custom Interface) * Examples: Standalone use of AWS Textract” Gemini 3.5 Flash · negative prompt · soft negative
“I would suggest starting with Google Cloud Document AI or Amazon Textract” MiniMax M2.5 · budget prompt · first choice
“If cost is the primary constraint, Amazon Textract is hard to beat” Claude Haiku 4.5 · budget prompt · first choice
“If you're already on AWS and need just extraction APIs, Textract's 3-month free tier and pay-per-page model is the cheapest to trial.” Muse Glimmer 30B · budget prompt · alternative
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.