Instructions to use Hcompany/Holo-3.1-35B-A3B-NVFP4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Hcompany/Holo-3.1-35B-A3B-NVFP4 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="Hcompany/Holo-3.1-35B-A3B-NVFP4")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("Hcompany/Holo-3.1-35B-A3B-NVFP4")
model = AutoModelForImageTextToText.from_pretrained("Hcompany/Holo-3.1-35B-A3B-NVFP4")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Hcompany/Holo-3.1-35B-A3B-NVFP4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Hcompany/Holo-3.1-35B-A3B-NVFP4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Hcompany/Holo-3.1-35B-A3B-NVFP4",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/Hcompany/Holo-3.1-35B-A3B-NVFP4

SGLang

How to use Hcompany/Holo-3.1-35B-A3B-NVFP4 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "Hcompany/Holo-3.1-35B-A3B-NVFP4" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Hcompany/Holo-3.1-35B-A3B-NVFP4",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "Hcompany/Holo-3.1-35B-A3B-NVFP4" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Hcompany/Holo-3.1-35B-A3B-NVFP4",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use Hcompany/Holo-3.1-35B-A3B-NVFP4 with Docker Model Runner:
```
docker model run hf.co/Hcompany/Holo-3.1-35B-A3B-NVFP4
```

Holo3.1: Fast & Local Computer Use Agents

This repository contains model weights in NVFP4 W4A16 format, obtained using Nvidia's Model-Optimizer library.

They were tested with vLLM 0.22.1rc1.dev112+g3f0a91bb9

Model Description

Holo3.1 is our latest family of Vision-Language Models (VLMs) for computer use agents. Building on Holo3, it expands support beyond browser and desktop automation to mobile environments, introduces native function-calling support for seamless integration with agent frameworks, and enables local deployment through optimized quantized checkpoints.

The Holo3.1 family spans model sizes from 0.8B to 35B-A3B parameters. Across computer use, UI grounding, mobile automation, and business workflows, Holo3.1 delivers strong performance while improving deployment flexibility and cost efficiency.

Developed by: H Company
Model type: Vision-Language Models for Navigation and Computer Use Agents
Available models: Holo3.1-0.8B, Holo3.1-4B, Holo3.1-9B, Holo3.1-35B-A3B
Base models: Qwen 3.5 family
Supported environments: Web, Desktop, Mobile
Available quantizations for Holo3.1-35B-A3B: BF16, FP8, NVFP4, Q4 GGUF
Blog Post: hcompany.ai/holo3.1
Quickstart: hub.hcompany.ai/quickstart
License: Apache 2.0 License

Performance vs Cost

The figure below compares the overall performance and inference cost of the Holo3.1 and Qwen 3.5 families. Overall performance averages computer use, mobile automation, enterprise workflows, and UI grounding benchmarks.

Holo3.1 establishes a strong Pareto frontier across model sizes, from lightweight local agents to state-of-the-art enterprise deployments.

Benchmark Results

Holo3.1 delivers strong performance across computer use, mobile automation, enterprise workflows, and UI grounding benchmarks.

Table 1: Evaluation results across computer use, mobile automation, enterprise workflows, and grounding benchmarks.

Get Started

Explore our Quickstart guide to learn how to integrate Holo3.1 into your applications, deploy local agents, or run optimized inference on NVIDIA hardware.

Citation

@misc{hai2026holo31,
      title={Holo3.1: Fast & Local Computer Use Agents},
      author={H Company},
      year={2026},
      url={https://huggingface.co/Hcompany/Holo3.1-35B-A3B},
}