Function Details

Chat Model

Provided by

Logo

OpenAI

Logo

Anthropic

Logo

TogetherAI

Logo

Google

Logo

Meta

Generate text using various Large Language Models (LLMs). This node supports multiple providers and models with different capabilities and pricing.

Available Models

Basic Models (Free)

- GPT-3.5 Turbo

- GPT-3.5 Turbo 16k

- Mistral (7B)

- LLaMa-2 Chat (70B)

- LLaMa-3 Chat (70B)

- GPT-4o Mini


Premium Models

- GPT-4 Turbo

- Claude 3 (Haiku, Sonnet, Opus)

- Gemini Pro

- Perplexity LLaMa 3.1

- Mixtral 8x7B


Key Features

  • Temperature control (0-1) for response creativity

  • System message support for model behavior control

  • Message history for conversation context

  • Image input support (GPT-4 Turbo, GPT-4o)

  • Forced JSON output option (GPT models only)

  • Citation control for Perplexity models

Citation Formats (Perplexity Models)

When using Perplexity models, you can control how citations are included in the response: 1. No Citations: Citations are completely omitted 2. Append to Text: Citations are added at the end of the response text 3. Separate JSON: Response is formatted as a JSON object:

{
    "content": "The AI-generated response text...",
    "citations": [
        "https://source1.com/article",
        "https://source2.com/reference",
        ...
    ]
}

Usage Tips

  • Lower temperature (0.1-0.3) for factual/consistent responses

  • Higher temperature (0.7-0.9) for creative content

  • Use system messages to define role and constraints

  • Chain responses using message history

  • For structured data, enable JSON output and specify format in prompt

  • When using Perplexity, choose citation format based on your needs:

  • Use "Append to Text" for human-readable output

  • Use "Separate JSON" for programmatic processing of citations

Example Usage

{
    "model": "perplexity/llama-3.1-sonar-large-128k-online",
    "text": "What are the latest developments in quantum computing?",
    "temperature": 0.3,
    "citation_format": "json"
}

{
    "model": "gpt-4o-mini",
    "text": "Write a product description for a coffee maker",
    "temperature": 0.7,
    "system-message": "You are a professional copywriter specializing in product descriptions."
}