Best Document Question Answering Models

Document question answering is the task of answering questions grounded in document images: invoices, forms, contracts, and reports. It layers three abilities: reading the text (OCR), understanding layout structure such as tables, key-value pairs, and reading order, and reasoning over both to locate the answer. Layout-aware transformers pioneered the recipe and modern vision language models now do it end to end from the page image. Benchmarks like DocVQA score answer accuracy. It automates data entry, claims and loan processing, compliance review, and contract analysis. This page lists 62 document question answering models, including 23 open-weight options you can self-host; all of them run live in the Playground so you can test them on your own images.

It depends on your task and constraints. For fixed categories in production, a model fine-tuned on your own data typically beats any general-purpose model. Compare the document question answering models on this page and try them on your own image to see which fits.

Yes. 23 of the 62 document question answering models here are open-weight (for example Kimi K3, Qwen3.6 27B, and Qwen3.6 35B A3B), free to self-host under their licenses (Modified MIT, Apache 2.0, and MIT).

Yes. You can run all 62 in the Roboflow Playground for free. Upload an image and compare the models' output side by side, no setup required.