LMM

Class: LMMBlockV1

Source: inference.core.workflows.core_steps.models.foundation.lmm.v1.LMMBlockV1

Ask a question to a Large Multimodal Model (LMM) with an image and text.

You can specify arbitrary text prompts to an LMMBlock.

The LLMBlock supports two LMMs:

• OpenAI's GPT-4 with Vision;

You need to provide your OpenAI API key to use the GPT-4 with Vision model.

If you want to classify an image into one or more categories, we recommend using the dedicated LMMForClassificationBlock.

Type identifier

Use the following identifier in step "type" field: roboflow_core/lmm@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	Holds unconstrained text prompt to LMM mode.	✅
lmm_type	str	Type of LMM to be used.	✅
lmm_config	LMMConfig	Configuration of LMM.	❌
remote_api_key	str	Holds API key required to call LMM model - in current state of development, we require OpenAI key when lmm_type=gpt_4v..	✅
json_output	Dict[str, str]	Holds dictionary that maps name of requested output field into its description.	❌

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 LMM in version v1.

• inputs: Multi-Label Classification Model, Single-Label Classification Model, Classification Label Visualization, Background Color Visualization, Webhook Sink, Dynamic Crop, Mask Visualization, Clip Comparison, Google Vision OCR, Twilio SMS Notification, Absolute Static Crop, Model Monitoring Inference Aggregator, Stability AI Image Generation, Florence-2 Model, Image Blur, LMM For Classification, Roboflow Dataset Upload, CogVLM, Circle Visualization, OCR Model, Crop Visualization, OpenAI, Stitch OCR Detections, OpenAI, Image Preprocessing, Model Comparison Visualization, Stitch Images, Bounding Box Visualization, Keypoint Detection Model, Perspective Correction, SIFT Comparison, Relative Static Crop, Slack Notification, Color Visualization, Ellipse Visualization, Reference Path Visualization, Blur Visualization, Pixelate Visualization, Anthropic Claude, Email Notification, LMM, Llama 3.2 Vision, CSV Formatter, VLM as Detector, Keypoint Visualization, Camera Focus, Florence-2 Model, Grid Visualization, Image Convert Grayscale, Image Threshold, Trace Visualization, Polygon Visualization, Triangle Visualization, Stability AI Inpainting, Halo Visualization, Dot Visualization, Polygon Zone Visualization, Google Gemini, Local File Sink, Instance Segmentation Model, Roboflow Custom Metadata, VLM as Classifier, Camera Calibration, Object Detection Model, SIFT, Corner Visualization, Image Contours, Line Counter Visualization, Roboflow Dataset Upload, Image Slicer, Image Slicer, Label Visualization
• outputs: Multi-Label Classification Model, Classification Label Visualization, Dimension Collapse, Background Color Visualization, Dynamic Crop, Clip Comparison, Segment Anything 2 Model, Absolute Static Crop, LMM For Classification, Image Blur, Roboflow Dataset Upload, CogVLM, Circle Visualization, OCR Model, Clip Comparison, Template Matching, SIFT Comparison, Multi-Label Classification Model, Path Deviation, OpenAI, Stitch OCR Detections, Detections Stitch, QR Code Detection, Pixel Color Count, Detections Stabilizer, Path Deviation, Line Counter, VLM as Classifier, Time in Zone, Model Comparison Visualization, Stitch Images, Bounding Box Visualization, Keypoint Detection Model, Moondream2, Continue If, Slack Notification, Color Visualization, LMM, Llama 3.2 Vision, Instance Segmentation Model, VLM as Detector, Time in Zone, Byte Tracker, Detections Filter, YOLO-World Model, Barcode Detection, Grid Visualization, SmolVLM2, Stability AI Inpainting, Keypoint Detection Model, Cosine Similarity, Dot Visualization, Google Gemini, Detections Merge, CLIP Embedding Model, Dynamic Zone, Size Measurement, Property Definition, Roboflow Custom Metadata, Data Aggregator, VLM as Classifier, Detections Classes Replacement, Gaze Detection, Object Detection Model, Corner Visualization, Roboflow Dataset Upload, Image Preprocessing, Image Slicer, Byte Tracker, Line Counter, Label Visualization, Identify Outliers, Bounding Rectangle, Single-Label Classification Model, First Non Empty Or Default, Webhook Sink, Cache Get, Mask Visualization, Delta Filter, Twilio SMS Notification, Google Vision OCR, Buffer, Detection Offset, Model Monitoring Inference Aggregator, Stability AI Image Generation, Florence-2 Model, Cache Set, Identify Changes, Crop Visualization, VLM as Detector, JSON Parser, Velocity, OpenAI, Dominant Color, Perspective Correction, SIFT Comparison, Relative Static Crop, Ellipse Visualization, Reference Path Visualization, Anthropic Claude, Blur Visualization, Pixelate Visualization, Email Notification, Expression, CSV Formatter, Keypoint Visualization, Camera Focus, Single-Label Classification Model, Qwen2.5-VL, Florence-2 Model, Detections Transformation, Trace Visualization, Image Threshold, Image Convert Grayscale, Detections Consensus, Triangle Visualization, Polygon Visualization, Halo Visualization, Polygon Zone Visualization, Local File Sink, Rate Limiter, Instance Segmentation Model, Camera Calibration, SIFT, Image Contours, Line Counter Visualization, Image Slicer, Byte Tracker, Distance Measurement, Object Detection Model

Input and Output Bindings

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

Bindings

• input

  • images (image): The image to infer on..
  • prompt (string): Holds unconstrained text prompt to LMM mode.
  • lmm_type (string): Type of LMM to be used.
  • remote_api_key (Union[string, secret]): Holds API key required to call LMM model - in current state of development, we require OpenAI key when lmm_type=gpt_4v..

• output

  • parent_id (parent_id): Identifier of parent for step output.
  • root_parent_id (parent_id): Identifier of parent for step output.
  • image (image_metadata): Dictionary with image metadata required by supervision.
  • structured_output (dictionary): Dictionary.
  • raw_output (string): String value.
  • * (*): Equivalent of any element.

Example JSON definition of step LMM in version v1

{
    "name": "<your_step_name_here>",
    "type": "roboflow_core/lmm@v1",
    "images": "$inputs.image",
    "prompt": "my prompt",
    "lmm_type": "gpt_4v",
    "lmm_config": {
        "gpt_image_detail": "low",
        "gpt_model_version": "gpt-4o",
        "max_tokens": 200
    },
    "remote_api_key": "xxx-xxx",
    "json_output": {
        "count": "number of cats in the picture"
    }
}