Instructions to use SL-AI/Openprose-2-Flash with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SL-AI/Openprose-2-Flash with Transformers:

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

pipe = pipeline("image-text-to-text", model="SL-AI/Openprose-2-Flash")
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, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("SL-AI/Openprose-2-Flash")
model = AutoModelForMultimodalLM.from_pretrained("SL-AI/Openprose-2-Flash", device_map="auto")
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 SL-AI/Openprose-2-Flash with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "SL-AI/Openprose-2-Flash"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SL-AI/Openprose-2-Flash",
		"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/SL-AI/Openprose-2-Flash

SGLang

How to use SL-AI/Openprose-2-Flash 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 "SL-AI/Openprose-2-Flash" \
    --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": "SL-AI/Openprose-2-Flash",
		"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 "SL-AI/Openprose-2-Flash" \
        --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": "SL-AI/Openprose-2-Flash",
		"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 SL-AI/Openprose-2-Flash with Docker Model Runner:
```
docker model run hf.co/SL-AI/Openprose-2-Flash
```

Openprose-2-Flash

File size: 2,463 Bytes

---
license: apache-2.0
language:
- en
- zh
- fr
- de
- es
- ja
- ko
- pt
- ru
- ar
pipeline_tag: image-text-to-text
library_name: transformers
base_model:
- SL-AI/GRaPE-2.1-Flash
tags:
- reasoning
- thinking_modes
- qwen3
- grape
- safetensors
- vision
- multimodal
- instruct
- chat
- coding
- math
- science
datasets:
- SL-AI/openprose
---

# The Openprose 2 Flash

| Model | Size | Modalities | Domain |
| :--- | :--- | :--- | :--- |
| **Openprose 2 Flash** | 9B | Image + Text in, Text out | On-device writing |

**Openprose 2 Flash** is a mid-sized model designed for creative writing tasks. Built on a **Qwen3.5** base, it supports multimodal inputs (image + text) and was trained on 200,000 entires of human-text for creative writing.

Openprose 2 Flash **does not** support any capability to think.

***

## What's New in Openprose 2

Openprose 2 Flash addresses several shortcomings from the first generation:

- **Stronger base model** — Built on Qwen3.5 9B, a substantially more capable foundation than the Qwen3 VL model used in the original model based on Openprose.
- **More time spent training** — With more tokens than the original model, it is capable of more complex ideas in story.
- **More parameters** — The 9B scale places Openprose 2 Flash boosts the parameter count of Openprose 1 by 6B, making it ever more smarter

***

# Capabilities

Openprose 2 Flash was post-trained on a curated dataset with heavy emphasis on creative writing, **it will perform poorly on other domains.**

GRaPE 2.1 Flash accepts **image and text** as input and produces **text** as output.

***

## Thinking Modes

Openprose 2 Flash **does not** support thinking. As we move to GRaPE 2.5, we will not only advance it's writing skills, but also support thinking.

***

# Recommended Inference Settings

Tested in **LM Studio**. These sampling parameters are a good starting point:

| Parameter | Value |
| :--- | :--- |
| **Temperature** | 0.6 |
| **Top K** | 20 |
| **Repeat Penalty** | 1.0 |
| **Top P** | 0.95 |
| **Min P** | 0 |

***

# Notes

- Openprose 2 Flash is the only Openprose 2 model that will be released.
- Training data is open-source.
- Updates and announcements are posted on [Skinnertopia](https://www.skinnertopia.com/) and this Hugging Face repository.

***

_Openprose 2 Flash is developed under the [SLAI (Skinnertopia Lab for Artificial Intelligence)](https://www.skinnertopia.com/) brand and released under the Apache 2.0 license._