Try
Reference

LanguageModelNode

Generates AI responses using language models from OpenAI, Groq, Anthropic, and more. This is the core AI node for generating conversational responses.

Overview

The LanguageModelNode is one of the most commonly used nodes in ConvoFlow. It takes a query (and optional context) and generates AI-powered responses using various language model services.

Inputs

query *

Type: string | Required: Yes

The main text query to send to the language model. Typically connected from QueryNode's output.

context

Type: string | Required: No

Additional context from knowledge base or other sources. This is often connected from KnowledgeBaseRetrievalNode's output to provide relevant information for the AI to use in its response.

Outputs

response

Type: string

The generated response from the language model. This is typically connected to ResponseNode for final output.

Parameters

service *

Type: string | Required: Yes

Language model service to use. Available options:

  • openai - OpenAI models (GPT-3.5, GPT-4, etc.)
  • groq - Groq models (fast inference)
  • ollama - Local Ollama models

Default: "openai"

model

Type: string | Required: No

Specific model to use. If left empty, the default model for the selected service will be used.

Default: "" (uses service default)

system_prompt

Type: string | Required: No

System/base prompt to set AI behavior. This defines the role and personality of the AI assistant.

Default: "You are a helpful AI assistant."

temperature

Type: float | Required: No

Creativity/randomness level. Range: 0.0 to 1.0. Lower values (0.0-0.5) produce more focused and consistent responses. Higher values (0.6-1.0) produce more creative and varied responses.

Default: 0.7

max_tokens

Type: integer | Required: No

Maximum number of tokens in the response. Controls the length of the generated text.

Default: 500

Tools Used

LanguageModelNode uses the LanguageModelTool to interact with language model APIs.

python
from tools.language_model_tool.language_model_tool import LanguageModelTool

tool = LanguageModelTool()
result = tool.generate_response(
    query="Hello, how are you?",
    service="openai",
    model="gpt-3.5-turbo",
    system_prompt="You are a helpful assistant.",
    temperature=0.7,
    max_tokens=500
)

Learn more about creating tools: Creating Tools

Required Credentials

Required API Keys:

  • OPENAI_API_KEY - Required for OpenAI service
  • GROQ_API_KEY - Required for Groq service
  • ANTHROPIC_API_KEY - Required for Anthropic service

Configure these in your environment variables or through the ConvoFlow admin interface.

Example Usage

Here's a common workflow pattern using LanguageModelNode:

javascript
// Basic AI Chatbot:
QueryNode (query: "What is AI?")
  ↓ [query]
LanguageModelNode (
  service: "openai",
  model: "gpt-3.5-turbo",
  system_prompt: "You are a helpful assistant.",
  temperature: 0.7
)
  ↓ [response]
ResponseNode
  → "AI, or Artificial Intelligence, is..."
javascript
// RAG (Retrieval-Augmented Generation):
QueryNode (query: "How do I install ConvoFlow?")
  ↓ [query]
KnowledgeBaseRetrievalNode (collection: "docs")
  ↓ [response as context]
LanguageModelNode (
  service: "openai",
  system_prompt: "Answer based on the provided context.",
  temperature: 0.3
)
  ↓ [response]
ResponseNode
  → "To install ConvoFlow, first..."

Styling

LanguageModelNode has a distinctive design to indicate it's an AI node:

  • Shape: Rounded rectangle
  • Border Color: Purple (#a78bfa)
  • Background: Dark (#1f1f1f)
  • Subtitle: "GENERATES TEXT"
  • Icon: Sparkles icon (✨)

Related Nodes

Was this page helpful?