SmolVLM2

Class: SmolVLM2BlockV1

Source: inference.core.workflows.core_steps.models.foundation.smolvlm.v1.SmolVLM2BlockV1

This workflow block runs SmolVLM2, a multimodal vision-language model. You can ask questions about images -- including documents and photos -- and get answers in natural language.

Type identifier

Use the following identifier in step "type" field: roboflow_core/smolvlm2@v1to add the block as as step in your workflow.

Properties

Name	Type	Description	Refs
name	str	Enter a unique identifier for this step..	❌
prompt	str	Optional text prompt to provide additional context to SmolVLM2. Otherwise it will just be None.	❌
model_version	str	The SmolVLM2 model to be used for inference..	✅

The Refs column marks possibility to parametrise the property with dynamic values available in workflow runtime. See Bindings for more info.

Available Connections

Compatible Blocks

Check what blocks you can connect to SmolVLM2 in version v1.

• inputs: Crop Visualization, SIFT, Stability AI Image Generation, Triangle Visualization, Blur Visualization, Background Color Visualization, Relative Static Crop, Color Visualization, Image Contours, Camera Focus, Corner Visualization, Line Counter Visualization, Icon Visualization, Mask Visualization, Image Convert Grayscale, Circle Visualization, Image Blur, Pixelate Visualization, Absolute Static Crop, Model Comparison Visualization, Multi-Label Classification Model, Image Threshold, Keypoint Detection Model, Reference Path Visualization, Image Slicer, Stitch Images, Depth Estimation, Trace Visualization, Image Preprocessing, Classification Label Visualization, Polygon Visualization, Stability AI Outpainting, Keypoint Visualization, Dot Visualization, Grid Visualization, Object Detection Model, Single-Label Classification Model, Bounding Box Visualization, Camera Calibration, Polygon Zone Visualization, Ellipse Visualization, QR Code Generator, Halo Visualization, Perspective Correction, Stability AI Inpainting, Image Slicer, Instance Segmentation Model, SIFT Comparison, Label Visualization, Dynamic Crop
• outputs: Detections Consensus

Input and Output Bindings

The available connections depend on its binding kinds. Check what binding kinds SmolVLM2 in version v1 has.

Bindings

• input

  • images (image): The image to infer on..
  • model_version (roboflow_model_id): The SmolVLM2 model to be used for inference..

• output

  • parsed_output (dictionary): Dictionary.

Example JSON definition of step SmolVLM2 in version v1

{
    "name": "<your_step_name_here>",
    "type": "roboflow_core/smolvlm2@v1",
    "images": "$inputs.image",
    "prompt": "What is in this image?",
    "model_version": "smolvlm2/smolvlm-2.2b-instruct"
}