ChatGoogleGenerativeAI

​

Node Details

• Name: chatGoogleGenerativeAI
• Type: ChatGoogleGenerativeAI
• Version: 2.1
• Category: Chat Models

​

Parameters

​

Parameters

​

Credential (Required)

​

Credential (Required)

• Type: googleGenerativeAI
• Description: Google Generative AI API key

​

Inputs

​

Inputs

1. Model Name (Required)
   • Type: asyncOptions
   • Default: “gemini-pro”
   • Description: The name of the Google Gemini model to use
2. Model Name (Required)
   • Type: asyncOptions
   • Default: “gemini-pro”
   • Description: The name of the Google Gemini model to use
3. Cache (Optional)
   • Type: BaseCache
   • Description: Caching mechanism for responses
4. Cache (Optional)
   • Type: BaseCache
   • Description: Caching mechanism for responses
5. Custom Model Name (Optional)
   • Type: string
   • Description: Override the selected model with a custom model name
6. Custom Model Name (Optional)
   • Type: string
   • Description: Override the selected model with a custom model name
7. Temperature (Optional)
   • Type: number
   • Default: 0.9
   • Range: 0 to 1
   • Description: Controls the randomness of the output
8. Temperature (Optional)
   • Type: number
   • Default: 0.9
   • Range: 0 to 1
   • Description: Controls the randomness of the output
9. Max Output Tokens (Optional)
   • Type: number
   • Description: Maximum number of tokens in the generated response
10. Max Output Tokens (Optional)
    • Type: number
    • Description: Maximum number of tokens in the generated response
11. Top Probability (Top P) (Optional)
    • Type: number
    • Range: 0 to 1
    • Description: Nucleus sampling, considers the smallest set of tokens whose cumulative probability exceeds the value
12. Top Probability (Top P) (Optional)
    • Type: number
    • Range: 0 to 1
    • Description: Nucleus sampling, considers the smallest set of tokens whose cumulative probability exceeds the value
13. Top Next Highest Probability Tokens (Top K) (Optional)
    • Type: number
    • Description: Considers only the top K most probable tokens for each step
14. Top Next Highest Probability Tokens (Top K) (Optional)
    • Type: number
    • Description: Considers only the top K most probable tokens for each step
15. Harm Category (Optional)
    • Type: multiOptions
    • Options: Dangerous, Harassment, Hate Speech, Sexually Explicit
    • Description: Categories of harmful content to filter
16. Harm Category (Optional)
    • Type: multiOptions
    • Options: Dangerous, Harassment, Hate Speech, Sexually Explicit
    • Description: Categories of harmful content to filter
17. Harm Block Threshold (Optional)
    • Type: multiOptions
    • Options: Low and Above, Medium and Above, None, Only High, Threshold Unspecified
    • Description: Threshold for blocking harmful content
18. Harm Block Threshold (Optional)
    • Type: multiOptions
    • Options: Low and Above, Medium and Above, None, Only High, Threshold Unspecified
    • Description: Threshold for blocking harmful content
19. Allow Image Uploads (Optional)
    • Type: boolean
    • Default: false
    • Description: Enables image upload capability, automatically using vision model when an image is uploaded
20. Allow Image Uploads (Optional)
    • Type: boolean
    • Default: false
    • Description: Enables image upload capability, automatically using vision model when an image is uploaded

​

Functionality

1. Initializes the ChatGoogleGenerativeAI model with the provided configuration
2. Applies safety settings based on selected harm categories and thresholds
3. Supports multi-modal interactions if image uploads are allowed
4. Integrates with various LangChain components like LLMChain, Conversation Chain, ReAct Agent, etc.

​

Usage

• Chatbots and conversational AI
• Content generation
• Text analysis and summarization
• Question-answering systems

​

Notes

• Ensure you have the necessary Google Generative AI API credentials before using this node.
• Be mindful of the model’s capabilities and limitations when designing your application.
• Properly configure safety settings to align with your application’s content policies.