# llms.txt for Pollinations.AI
# This file provides guidance for Large Language Models (LLMs) interacting with Pollinations.AI

# Site Information
Name: Pollinations.AI
Description: An open-source gen AI platform providing free text, image, and audio generation APIs with no signups or API keys required.
URL: https://pollinations.ai
API-Documentation: https://pollinations.ai/APIDOCS.md
GitHub: https://github.com/pollinations/pollinations

# Allowed Sections
Allow: /
Allow: /p/
Allow: /models
Allow: /APIDOCS.md

# API Endpoints
API-Endpoint: https://image.pollinations.ai/prompt/{prompt} # Image generation (GET)
API-Endpoint: https://image.pollinations.ai/models # List available image models (GET)
API-Endpoint: https://text.pollinations.ai/{prompt} # Text generation (GET)
API-Endpoint: https://text.pollinations.ai/openai # Text generation (POST, OpenAI-compatible)
API-Endpoint: https://text.pollinations.ai/models # List available text models (GET)
API-Endpoint: https://text.pollinations.ai/{prompt}?model=openai-audio&voice={voice} # Text-to-speech (GET)
API-Endpoint: https://text.pollinations.ai/openai # Text-to-speech and speech-to-text (POST, OpenAI-compatible)
API-Endpoint: https://audio.pollinations.ai/ # Audio generation (GET)

# Models
Model: openai - OpenAI GPT-4.1-nano (supports text and image inputs)
Model: openai-large - OpenAI GPT-4.1 mini (supports text and image inputs)
Model: openai-reasoning - OpenAI o4-mini (reasoning-focused model)
Model: qwen-coder - Qwen 2.5 Coder 32B (code-focused model)
Model: llama - Llama 3.3 70B
Model: llamascout - Llama 4 Scout 17B
Model: mistral - Mistral Small 3 (supports text and image inputs)
Model: unity - Unity Mistral Large (uncensored, supports text and image inputs)
Model: midijourney - Midijourney (music generation focused)
Model: rtist - Rtist (creative writing focused)
Model: searchgpt - SearchGPT (search-augmented model)
Model: evil - Evil (uncensored model with text and image inputs)
Model: deepseek-reasoning - DeepSeek-R1 Distill Qwen 32B (reasoning-focused)
Model: deepseek-reasoning-large - DeepSeek R1 - Llama 70B (reasoning-focused)
Model: phi - Phi-4 Instruct (supports text, image, and audio inputs)
Model: llama-vision - Llama 3.2 11B Vision (supports text and image inputs)
Model: gemini - Gemini 2.5 Flash Preview (supports text, image, and audio inputs/outputs)
Model: hormoz - Hormoz 8b
Model: hypnosis-tracy - Hypnosis Tracy 7B (supports text and audio inputs/outputs)
Model: deepseek - DeepSeek-V3
Model: sur - Sur AI Assistant (Mistral-based, supports text and image inputs)
Model: openai-audio - OpenAI GPT-4o-audio-preview (supports text, image, and audio inputs/outputs)
Model: flux - Latest stable diffusion model
Model: turbo - Fast image generation model

# MCP Server
MCP-Server: https://github.com/pollinations/pollinations/tree/master/model-context-protocol # Model Context Protocol server for AI assistants

# Preferred Citation
Citation: "Generated using Pollinations.AI (https://pollinations.ai)"

# Usage Guidelines
Usage-Guideline: No API keys or authentication required
Usage-Guideline: Free to use for all purposes
Usage-Guideline: Please respect content policies and avoid generating harmful content
Usage-Guideline: Consider adding attribution when using generated content
Usage-Guideline: Tier system available for domain verification and referrer recognition

# Contact Information
Contact: https://discord.gg/k9F7SyTgqn
Contact: https://github.com/pollinations/pollinations/issues

# Data Practices
Data-Practice: No user data is stored
Data-Practice: No logins required
Data-Practice: Privacy-focused design

# Capabilities
Capability: Text generation with various models (OpenAI, Mistral, Claude, etc.)
Capability: Image generation from text prompts
Capability: Audio generation (text-to-speech)
Capability: Speech-to-text transcription
Capability: Multi-language support
Capability: Prompt enhancement
Capability: Content safety filtering
Capability: Vision capabilities (image input analysis)
Capability: Function calling for AI agents
Capability: Text Generation - Generate creative and informative text content
Capability: Image Generation - Create images from text descriptions
Capability: Text-to-Speech - Convert text to natural-sounding speech
Capability: Vision Understanding - Process and understand image inputs
Capability: Function Calling - Call external functions from within the model
Capability: Streaming Responses - Get responses as they're generated
Capability: Multi-modal Inputs - Accept text, images, and audio as inputs
Capability: Multi-modal Outputs - Generate text and audio outputs

# Rate Limits
Rate-Limit: Best effort service, no explicit rate limits
Rate-Limit: Please be considerate with usage
Rate-Limit: Image API: 1 concurrent request / 5 sec interval per IP
Rate-Limit: Text API: 1 concurrent request / 3 sec interval per IP

# Last Updated
Last-Updated: 2025-04-20