nvidia/PixelUMM

image-to-text model by nvidia

Open original source ↗

Older snapshot: check the original source for current information.

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.

PixelUMM: Encoder-Free Unified Image and Video Understanding and Generation Paper: PixelUMM: Encoder-Free Unified Image and Video Understanding and Generation Project: PixelUMM Project Page Code: GitHub PixelUMM is an NVIDIA-developed, encoder-free unified multimodal model for joint understanding and generation across text, images, and video directly in pixel space. Images are represented as pixel patches instead of embeddings from a separate pretrained vision encoder, allowing understanding and generation to share a Transformer-based representation. PixelUMM supports: an iterative pixel-generation head. Qwen/Qwen3-8B, revision b968826d9c46dd6066d109eabc6255188de91218. PixelUMM accepts text, RGB images, and video frames. It produces text, RGB images, and video frames. PixelUMM requires Linux, an NVIDIA GPU, a CUDA build of PyTorch, and FlashAttention. Install PyTorch and FlashAttention for your CUDA environment, then install the remaining dependencies. PixelUMM is intended for researchers and developers studying unified multimodal modeling, pixel-space representation learning, multimodal understanding, and image or video generation. Example uses include: PixelUMM has not been…

nvidia/PixelUMM on Hugging Face A short preview, not the full document.

Read the full model card ↗ · Preview checked 2026-11-19T08:41:32.483Z

Inside the original model card — Document outline
  1. PixelUMM
  2. Overview
  3. Model Architecture
  4. Inputs and Outputs
  5. Installation
  6. Intended Use
  7. Limitations and Safety
  8. License
  9. References

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

Links from the model card

References supplied by the publisher, not independently verified endorsements. Check the destination before downloading files or entering credentials.

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

nvidia/PixelUMM is a image-to-text model repository published by nvidia 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-to-text
Library
Not reported
Recent downloads (30 days)
0
Cumulative likes
81
Hugging Face trending score
50
Reported safetensors parameters
Not reported
Architecture
Not reported
License
other
Access
Not gated by Hugging Face
Created
2026-10-01T21:40:34.000Z
Last modified
2026-10-01T21:29:11.000Z

Compatibility and lineage

Language coverage was not reported.

Reported base models: Qwen/Qwen3-8B

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 2026-10-11T06:40:09.055Z. 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

datalab-to/indra-ocr

iapp/OpenThai2.1-OCR-Text-20261118

iapp/OpenThai2.0-OCR-Text

Artificial-Soul/Devika

INGMUYLEANG/chakkhu-ocr

Litap007/indic-citizen-information-board-pmgsy-florence2-lora