iapp/OpenThai2.1-OCR-Markdown-20261118

image-text-to-text model by iapp

Open original source ↗

From the publisher

Model card & documentation

Source preview

Read the publisher’s intended use, setup instructions, evaluations and limitations. The original model card is the source of truth.

Thai and English document pages to Markdown, with tables as HTML. It is a Thai fine-tune of NVIDIA Nemotron 3.5 Parse (0.63B) from OpenThai (formerly OpenThaiGPT). Developed by iApp, launched with AIEAT. It keeps Parse's layout boxes, block classes and reading order, and its whole serving and postprocessing stack, so it works anywhere Parse does. New in 2.2: DocumentParser(..., dewarp=True, dewarpmodels="dewarp") flattens bent or tilted phone photos before they are sent to the model (see Use). The weights and every score below are 2.1's. In our dev tests it made no difference beyond run-to-run variation for this model, which was trained on photographed pages, so it is off and not recommended yet; it does help OpenThai2.1-OCR-Text on phone photos. New in 2.1 (against OpenThai2.0-OCR-Markdown, tag v2.0.0): government documents: academic papers, books, comics, exams, magazines, newspapers, reports, slides and textbooks, in Thai and English. 80.1). $\begin{gathered}...\end{gathered}$. text on it. colspan="1". tables…

iapp/OpenThai2.1-OCR-Markdown-20261118 on Hugging Face A short preview, not the full document.

Read the full model card ↗ · Preview checked 2026-11-18T09:45:29.705Z

Inside the original model card — Document outline
  1. OpenThai2.1-OCR-Markdown-20261118
  2. Model at a glance
  3. Results
  4. Use
  5. Training
  6. Limits
  7. Acknowledgements
  8. Licence

Headings are captured from the source. Links open the publisher’s document, not a locally hosted copy.

Documentation belongs to its respective authors. Reported project/model license: other. A listing is not a grant of reuse or training rights. Confirm the document’s own terms at the source.

Model overview

iapp/OpenThai2.1-OCR-Markdown-20261118 is a image-text-to-text model repository published by iapp on Hugging Face. A library was not reported.

This page summarizes Hub metadata. For intended use, training data, evaluation results and limitations, consult the original model card.

Model facts

Task
image-text-to-text
Library
Not reported
Recent downloads (30 days)
12,448
Cumulative likes
52
Hugging Face trending score
38
Reported safetensors parameters
630,800,504
Architecture
RadioMBartPrefixForCausalLM
License
other
Access
Not gated by Hugging Face
Created
2026-11-09T14:21:29.000Z
Last modified
2026-11-18T04:59:47.000Z

Compatibility and lineage

Reported languages: th, en

Reported base models: nvidia/Nemotron-3.5-Parse

Parameter count is not a RAM/VRAM requirement. Check precision, quantization, context length and runtime compatibility in the model card; no hardware or API-cost claim is inferred here.

Use and evaluate

Check license terms and access requirements first. Review the model card’s documented loading instructions, evaluations and safety limitations. Benchmark scores are not imported by this directory, and popularity is not an accuracy ranking.

No model weights are downloaded or executed by AltAPIs. Never enable remote model code without reviewing it.

Model card and usage instructions

Model files and configuration

Community discussion

Source and freshness

Source: Hugging Face Hub. Metadata observed 2026-11-18T06:59:26.317Z. Daily imports are snapshots, not real-time monitoring.

Popularity and source listings do not establish security, suitability, licensing rights or benchmark performance.

Open original source

Related models

microsoft/Florence-2-base

ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-GGUF

Qwen/Qwen3.8-27B

Qwen/Qwen3-VL-8B-Instruct

zai-org/GLM-5.3-Flash

google/gemma-4-31B-it