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

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"
}