Qwen3.5

Use Alibaba's Qwen3.5-VL vision-language model on a Dedicated Deployment or self-hosted Inference

Qwen3.5 is Alibaba's vision-language model family. It accepts an image and a text prompt and returns a text response. We support two pretrained checkpoints in Inference:

Alias	Parameters
qwen3_5-0.8b	0.8B
qwen3_5-2b	2B

Qwen3.5 is not available on the Serverless Hosted API. Run it on a Dedicated Deployment or self-hosted Inference.

Code sample

Install the Inference SDK:

pip install inference-sdk

Set api_url to your Dedicated Deployment URL or a local Inference server. Pass your Roboflow API Key via the API_KEY environment variable.

import os
import urllib.request
from inference_sdk import InferenceHTTPClient

image_url = "https://media.roboflow.com/notebooks/examples/dog.jpeg"
image_path = "dog.jpeg"
urllib.request.urlretrieve(image_url, image_path)

client = InferenceHTTPClient(
    api_url="https://your-deployment.roboflow.cloud",
    api_key=os.getenv("API_KEY"),
)
result = client.infer_lmm(
    image_path,
    model_id="qwen3_5-2b",
    prompt="Describe this image briefly.",
    max_new_tokens=128,
)
print(result["response"])

The code above prints the model response to the terminal:

This is a close-up, eye-level shot of a blue beagle dog being carried on the shoulders of a person. The dog has its tongue lolling out in a relaxed, happy expression, looking slightly upward and to the right. Its fur is a vibrant, solid blue with white markings on its chest and muzzle.

Set api_url to match your deployment target:

• http://localhost:9001 for a local Inference server.

• Your Dedicated Deployment URL for a private endpoint.

You can train your own Qwen3.5 checkpoint on Roboflow and call it by its workspace/project/version identifier.