Pollinations.AI Text icon

Pollinations.AI Text

Generate text using Pollinations.AI
Join our community

Actions3

Overview

This node integrates with Pollinations.AI to generate text using advanced AI models. It supports generating text from conversation messages with customizable parameters such as model selection, creativity (temperature), maximum response length, reasoning depth, and streaming output. This node is useful for applications requiring sophisticated AI-driven text generation, such as chatbots, content creation, and interactive assistants.

Use Case Examples

  1. Generating a detailed AI response based on a conversation history with system, user, and assistant roles.
  2. Creating creative text outputs with controlled creativity and reasoning effort.
  3. Streaming AI-generated text responses in real-time for interactive applications.

Properties

Name Meaning
Messages Conversation messages consisting of multiple message entries with roles (system, user, assistant) and content.
Model The AI model to use for text generation.
Temperature Controls creativity of the AI response, from strict (0.0) to wild (3.0).
Max Tokens Maximum length of the AI-generated response in tokens.
Reasoning Effort How deeply the AI thinks before responding, ranging from minimal to high.
Stream Whether to receive the AI response in real-time chunks (streaming).

Output

JSON

  • model - The AI model used for generating the response.
  • response - The generated text response from the AI.
  • fullResponse - The full raw response object from the AI API, including additional metadata.

Dependencies

  • Pollinations.AI API with an API token credential

Troubleshooting

  • Missing or invalid API token will cause authentication errors; ensure the Pollinations.AI API token is correctly set in credentials.
  • If no messages are provided for the advanced generation operation, an error is thrown indicating that at least one message is required.
  • Timeouts or network errors may occur; verify network connectivity and API availability.
  • Errors from the API are returned in the output if 'continue on fail' is enabled, otherwise they cause the node to fail.

Links

Discussion