Google Gemma API¶
Deprecated
Use the Google Gemma v2 block, which adds a Roboflow-managed API key option, user-selectable privacy controls, and the same OpenRouter passthrough capabilities.
Class: GoogleGemmaBlockV1
Source: inference.core.workflows.core_steps.models.foundation.google_gemma.v1.GoogleGemmaBlockV1
Ask a question to Google's Gemma model with vision capabilities.
You can specify arbitrary text prompts or predefined ones, the block supports the following types of prompt:
-
Open Prompt (
unconstrained) - Use any prompt to generate a raw response -
Text Recognition (OCR) (
ocr) - Model recognizes text in the image -
Visual Question Answering (
visual-question-answering) - Model answers the question you submit in the prompt -
Captioning (short) (
caption) - Model provides a short description of the image -
Captioning (
detailed-caption) - Model provides a long description of the image -
Single-Label Classification (
classification) - Model classifies the image content as one of the provided classes -
Multi-Label Classification (
multi-label-classification) - Model classifies the image content as one or more of the provided classes -
Unprompted Object Detection (
object-detection) - Model detects and returns the bounding boxes for prominent objects in the image -
Structured Output Generation (
structured-answering) - Model returns a JSON response with the specified fields
🛠️ API providers and model variants¶
Gemma is exposed via OpenRouter API and we require passing an OpenRouter API Key to run.
Pick a specific model version from the model_version dropdown - new Gemma releases
will be added to this list as they become available on OpenRouter.
API Usage Charges
OpenRouter is an external third party providing access to the model and incurring charges on the usage. Please check pricing on openrouter.ai before use.
💡 Further reading and Acceptable Use Policy¶
Model license
Check the Gemma Terms of Use before use.
Type identifier¶
Use the following identifier in step "type" field: roboflow_core/google_gemma@v1to add the block as
as step in your workflow.
Properties¶
| Name | Type | Description | Refs |
|---|---|---|---|
name |
str |
Enter a unique identifier for this step.. | ❌ |
task_type |
str |
Task type to be performed by model. Value determines required parameters and output response.. | ❌ |
prompt |
str |
Text prompt to the Gemma model. | ✅ |
output_structure |
Dict[str, str] |
Dictionary with structure of expected JSON response. | ❌ |
classes |
List[str] |
List of classes to be used. | ✅ |
api_key |
str |
Your OpenRouter API key. | ✅ |
model_version |
str |
Model to be used. | ✅ |
max_tokens |
int |
Maximum number of tokens the model can generate in it's response.. | ❌ |
temperature |
float |
Temperature to sample from the model - value in range 0.0-2.0, the higher - the more random / "creative" the generations are.. | ✅ |
max_concurrent_requests |
int |
Number of concurrent requests that can be executed by block when batch of input images provided. If not given - block defaults to value configured globally in Workflows Execution Engine. Please restrict if you hit limits.. | ❌ |
The Refs column marks possibility to parametrise the property with dynamic values available
in workflow runtime. See Bindings for more info.
Runtime compatibility¶
-
requires_internet— air-gapped / offline deployments - This block depends on a service that is not reachable from fully offline / air-gapped deployments.
Available Connections¶
Compatible Blocks
Check what blocks you can connect to Google Gemma API in version v1.
- inputs:
VLM As Classifier,MoonshotAI Kimi,Stability AI Image Generation,Trace Visualization,Image Stack,Anthropic Claude,Icon Visualization,SIFT Comparison,Morphological Transformation,Color Visualization,LMM For Classification,Perspective Correction,Corner Visualization,Clip Comparison,Roboflow Custom Metadata,Halo Visualization,Dynamic Zone,Qwen-VL,Email Notification,Halo Visualization,Google Gemma,Background Color Visualization,Ellipse Visualization,Email Notification,Twilio SMS/MMS Notification,Text Display,Polygon Visualization,Crop Visualization,Absolute Static Crop,Image Preprocessing,Model Monitoring Inference Aggregator,Relative Static Crop,OpenRouter,OpenAI,PLC ModbusTCP,Florence-2 Model,OpenAI,Heatmap Visualization,OCR Model,Motion Detection,Blur Visualization,Dimension Collapse,Depth Estimation,Instance Segmentation Model,Stability AI Outpainting,Anthropic Claude,Google Gemini,Clip Comparison,Google Gemini,PLC EthernetIP,Background Subtraction,Keypoint Visualization,CSV Formatter,Buffer,Webhook Sink,Stitch Images,Florence-2 Model,Current Time,Detections List Roll-Up,Contrast Equalization,OpenAI,VLM As Detector,Google Gemini,Triangle Visualization,Slack Notification,SIFT,Local File Sink,Cosine Similarity,Image Contours,Keypoint Detection Model,GLM-OCR,Roboflow Asset Library Attributes,Image Slicer,Polygon Zone Visualization,Contrast Enhancement,Google Gemma API,Stitch OCR Detections,Image Threshold,Line Counter Visualization,Camera Calibration,QR Code Generator,S3 Sink,Microsoft SQL Server Sink,Google Vision OCR,Twilio SMS Notification,Image Blur,Morphological Transformation,Camera Focus,Roboflow Vision Events,Size Measurement,Stability AI Inpainting,Classification Label Visualization,Stitch OCR Detections,Event Writer,Grid Visualization,Qwen3.5-VL,Mask Visualization,Llama 3.2 Vision,Reference Path Visualization,Image Slicer,Label Visualization,OPC UA Writer Sink,Dot Visualization,Identify Changes,Dynamic Crop,Circle Visualization,Llama 3.2 Vision,Camera Focus,Gaze Detection,OpenAI-Compatible LLM,MoonshotAI Kimi,Single-Label Classification Model,CogVLM,Qwen 3.6 API,Bounding Box Visualization,Multi-Label Classification Model,LMM,OpenAI,Image Convert Grayscale,Roboflow Visual Search,EasyOCR,Roboflow Dataset Upload,Pixelate Visualization,Roboflow Dataset Upload,PLC Writer,Qwen 3.5 API,Anthropic Claude,Object Detection Model,MQTT Writer,Polygon Visualization,Model Comparison Visualization - outputs:
VLM As Classifier,Line Counter,MoonshotAI Kimi,Stability AI Image Generation,Trace Visualization,Path Deviation,Anthropic Claude,Icon Visualization,SIFT Comparison,Morphological Transformation,Color Visualization,LMM For Classification,Perspective Correction,Corner Visualization,Clip Comparison,Roboflow Custom Metadata,Halo Visualization,Qwen-VL,Keypoint Detection Model,JSON Parser,Email Notification,Halo Visualization,Object Detection Model,Google Gemma,Background Color Visualization,Email Notification,Ellipse Visualization,Twilio SMS/MMS Notification,Text Display,Polygon Visualization,Crop Visualization,Image Preprocessing,Model Monitoring Inference Aggregator,OpenRouter,OpenAI,Florence-2 Model,VLM As Detector,OpenAI,Heatmap Visualization,Motion Detection,Perception Encoder Embedding Model,Depth Estimation,Instance Segmentation Model,Stability AI Outpainting,Anthropic Claude,YOLO-World Model,Google Gemini,Clip Comparison,Google Gemini,PLC EthernetIP,Keypoint Visualization,Buffer,Webhook Sink,Florence-2 Model,Current Time,Detections List Roll-Up,Contrast Equalization,OpenAI,Moondream2,Line Counter,VLM As Detector,Google Gemini,Triangle Visualization,Slack Notification,Time in Zone,CLIP Embedding Model,Multi-Label Classification Model,Local File Sink,VLM As Classifier,Keypoint Detection Model,Pixel Color Count,GLM-OCR,Roboflow Asset Library Attributes,Polygon Zone Visualization,Google Gemma API,Time in Zone,Stitch OCR Detections,Line Counter Visualization,Semantic Segmentation Model,Distance Measurement,Image Threshold,QR Code Generator,S3 Sink,Microsoft SQL Server Sink,Twilio SMS Notification,Google Vision OCR,Image Blur,Morphological Transformation,Roboflow Vision Events,Size Measurement,PTZ Tracking (ONVIF),Stability AI Inpainting,Classification Label Visualization,Stitch OCR Detections,Event Writer,Qwen3.5-VL,Grid Visualization,Llama 3.2 Vision,Mask Visualization,Reference Path Visualization,Label Visualization,OPC UA Writer Sink,Dot Visualization,Cache Set,Dynamic Crop,Detections Stitch,Circle Visualization,Llama 3.2 Vision,Path Deviation,SAM3 Video Tracker,Segment Anything 2 Model,OpenAI-Compatible LLM,MoonshotAI Kimi,CogVLM,Object Detection Model,Qwen 3.6 API,Detections Consensus,Bounding Box Visualization,LMM,OpenAI,SAM 3,PLC Reader,Instance Segmentation Model,Roboflow Visual Search,Roboflow Dataset Upload,SAM 3,Cache Get,Instance Segmentation Model,Detections Classes Replacement,Keypoint Detection Model,Instance Segmentation Model,Roboflow Dataset Upload,Qwen 3.5 API,Anthropic Claude,Object Detection Model,Time in Zone,MQTT Writer,Polygon Visualization,SAM 3,Model Comparison Visualization,Single-Label Classification Model,Seg Preview
Input and Output Bindings¶
The available connections depend on its binding kinds. Check what binding kinds
Google Gemma API in version v1 has.
Bindings
-
input
images(image): The image to infer on..prompt(string): Text prompt to the Gemma model.classes(list_of_values): List of classes to be used.api_key(Union[secret,string]): Your OpenRouter API key.model_version(string): Model to be used.temperature(float): Temperature to sample from the model - value in range 0.0-2.0, the higher - the more random / "creative" the generations are..
-
output
output(Union[string,language_model_output]): String value ifstringor LLM / VLM output iflanguage_model_output.classes(list_of_values): List of values of any type.
Example JSON definition of step Google Gemma API in version v1
{
"name": "<your_step_name_here>",
"type": "roboflow_core/google_gemma@v1",
"images": "$inputs.image",
"task_type": "<block_does_not_provide_example>",
"prompt": "my prompt",
"output_structure": {
"my_key": "description"
},
"classes": [
"class-a",
"class-b"
],
"api_key": "xxx-xxx",
"model_version": "Gemma 4 31B - OpenRouter",
"max_tokens": "<block_does_not_provide_example>",
"temperature": "<block_does_not_provide_example>",
"max_concurrent_requests": "<block_does_not_provide_example>"
}