Moondream2

Class: Moondream2BlockV1

Source: inference.core.workflows.core_steps.models.foundation.moondream2.v1.Moondream2BlockV1

This workflow block runs Moondream2, a multimodal vision-language model. You can use this block to run zero-shot object detection.

Type identifier

Use the following identifier in step "type" field: roboflow_core/moondream2@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 Moondream2. Otherwise it will just be None.	❌
model_version	str	The Moondream2 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 Moondream2 in version v1.

• inputs: Keypoint Visualization, Multi-Label Classification Model, Background Color Visualization, Polygon Zone Visualization, Stitch Images, Instance Segmentation Model, Polygon Visualization, Crop Visualization, Image Convert Grayscale, Grid Visualization, Image Blur, Mask Visualization, Relative Static Crop, SIFT, Camera Calibration, Circle Visualization, Image Slicer, Color Visualization, Image Slicer, Label Visualization, Dot Visualization, Absolute Static Crop, Blur Visualization, Pixelate Visualization, Dynamic Crop, Corner Visualization, Single-Label Classification Model, Image Preprocessing, SIFT Comparison, Image Threshold, Stability AI Image Generation, Trace Visualization, Bounding Box Visualization, Camera Focus, Ellipse Visualization, Stability AI Inpainting, Perspective Correction, Model Comparison Visualization, Depth Estimation, Reference Path Visualization, Line Counter Visualization, Triangle Visualization, Object Detection Model, Halo Visualization, Image Contours, Classification Label Visualization, Keypoint Detection Model
• outputs: Detections Stabilizer, Background Color Visualization, Stitch OCR Detections, Size Measurement, Florence-2 Model, Florence-2 Model, Detections Stitch, Crop Visualization, Path Deviation, Byte Tracker, Distance Measurement, Line Counter, Time in Zone, Detections Merge, Path Deviation, Circle Visualization, Color Visualization, Detection Offset, Detections Classes Replacement, Label Visualization, Dot Visualization, Blur Visualization, Pixelate Visualization, Dynamic Crop, Roboflow Dataset Upload, Corner Visualization, Detections Consensus, Time in Zone, Byte Tracker, Trace Visualization, Velocity, Bounding Box Visualization, Byte Tracker, Ellipse Visualization, Roboflow Custom Metadata, Model Monitoring Inference Aggregator, Roboflow Dataset Upload, Line Counter, Model Comparison Visualization, Perspective Correction, Overlap Filter, Detections Filter, Triangle Visualization, Detections Transformation, Segment Anything 2 Model

Input and Output Bindings

The available connections depend on its binding kinds. Check what binding kinds Moondream2 in version v1 has.

Bindings

• input

  • images (image): The image to infer on..
  • model_version (roboflow_model_id): The Moondream2 model to be used for inference..

• output

  • predictions (object_detection_prediction): Prediction with detected bounding boxes in form of sv.Detections(...) object.

Example JSON definition of step Moondream2 in version v1

{
    "name": "<your_step_name_here>",
    "type": "roboflow_core/moondream2@v1",
    "images": "$inputs.image",
    "prompt": "What is in this image?",
    "model_version": "moondream2/moondream2-2b"
}