Instructions to use nvidia/Nemotron-3.5-Parse with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use nvidia/Nemotron-3.5-Parse with Transformers:

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

pipe = pipeline("image-text-to-text", model="nvidia/Nemotron-3.5-Parse", trust_remote_code=True)
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)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("nvidia/Nemotron-3.5-Parse", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use nvidia/Nemotron-3.5-Parse with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "nvidia/Nemotron-3.5-Parse"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "nvidia/Nemotron-3.5-Parse",
		"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/nvidia/Nemotron-3.5-Parse

SGLang

How to use nvidia/Nemotron-3.5-Parse 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 "nvidia/Nemotron-3.5-Parse" \
    --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": "nvidia/Nemotron-3.5-Parse",
		"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 "nvidia/Nemotron-3.5-Parse" \
        --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": "nvidia/Nemotron-3.5-Parse",
		"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 nvidia/Nemotron-3.5-Parse with Docker Model Runner:
```
docker model run hf.co/nvidia/Nemotron-3.5-Parse
```

Nemotron-3.5-Parse

File size: 3,072 Bytes

{
  "status": "scored_and_verified",
  "scope": "Fresh quick20/full500 predictions and released Table-body conversion",
  "model": "nvidia/Nemotron-3.5-Parse",
  "model_revision": "dca5b2df89d00a671509204501d8f1f379d5e94f",
  "model_safetensors_sha256": "81dd7609c589f3694ce75520c25af5010953231358de6aec4d03e8c6aac787e9",
  "package_version": "3.5.0.post7",
  "scheduler": {
    "job_id": 7482504,
    "state": "COMPLETED",
    "exit_code": "0:0"
  },
  "runtime": {
    "gpu": "NVIDIA H100 80GB HBM3",
    "gpu_count": 1,
    "dtype": "bfloat16",
    "vllm": "0.26.0",
    "transformers": "5.14.1",
    "huggingface_hub": "1.24.0",
    "pillow": "12.3.0",
    "dependency_lock_reproduction": false,
    "deviation": "Measured serving container used Pillow 12.3.0; source package pins 12.2.0."
  },
  "generation": {
    "prompt": "</s><s><predict_bbox><predict_classes><output_markdown><predict_text_in_pic>",
    "concurrency": 128,
    "warmup_count": 16,
    "preload_images": true,
    "temperature": 0,
    "repetition_penalty": 1.1,
    "max_tokens": 8744,
    "max_model_len": 16384,
    "max_num_seqs": 192,
    "max_num_batched_tokens": 16384,
    "prefix_caching": false,
    "one_token_mtp": true,
    "attention": "FA3, automatic compatible scheduler",
    "coordinate_frame": "content_normalized"
  },
  "scoring": {
    "version": 3,
    "table_mode": "released",
    "conversion": "nemotron_parse.postprocessing.postprocess_text(Table,table_format=HTML,PostprocessOptions(render_charts=False))",
    "ground_truth_changed": false,
    "geometry_changed": false,
    "non_table_blocks_changed": false,
    "malformed_outputs": "retained"
  },
  "regression": {
    "quick20": {
      "pages": 20,
      "request_errors": 0,
      "finish_reasons": {
        "length": 1,
        "stop": 19
      },
      "table_conversion": {
        "changed_blocks": 13,
        "pages_with_predicted_tables": 10,
        "table_blocks": 14,
        "unconverted_latex_blocks": 1
      }
    },
    "full500": {
      "pages": 500,
      "request_errors": 0,
      "finish_reasons": {
        "length": 19,
        "stop": 481
      },
      "table_conversion": {
        "changed_blocks": 364,
        "pages_with_predicted_tables": 255,
        "table_blocks": 375,
        "unconverted_latex_blocks": 2
      }
    }
  },
  "verification": [
    {
      "split": "quick20",
      "pages": 20,
      "matches_measured_metrics_and_every_page": true,
      "tolerance": 1e-12
    },
    {
      "split": "full500",
      "pages": 500,
      "matches_measured_metrics_and_every_page": true,
      "tolerance": 1e-12
    }
  ],
  "limitations": [
    "Full500 is a lightweight regression harness, not official OmniDocBench TEDS or ParseBench.",
    "Token-capped predictions and 2 unconverted malformed Table bodies are retained.",
    "Timing excludes preload, warmup, startup, conversion, and scoring.",
    "Historical inference.constraints.txt and transformers.constraints.txt describe the older rc3 validation; they are not locks for this measured environment."
  ]
}