Pollinations: free big model services in the form of URL splicing and APIs

General Introduction

Pollinations is a fully open-source platform developed by the Berlin-based Pollination.AI team that provides free image, text and audio generation services. Users don't need to register or request an API key to use it via the web or API. It supports a wide range of AI models, including Flux Image generation, OpenAI text generation and audio conversion. The platform focuses on privacy protection, does not store any user data, and is anonymized throughout its use. To date, it serves more than 50,000 active users and generates more than 20 million images per month, which are widely used in the fields of creativity, development and education.

Pollinations was first a free image generation tool that allowed large models to generate articles with accompanying images. It works by splicing the prompt into the URL to generate an image: https://pollinations.ai/p/{prompt}

Pollinations has a while to turn its focus to AI music video creation and an exciting real-time immersive AI product called Dreamachine. for which I've created alternatives: https://img.kdjingpai.com/{prompt}

Recently, I noticed that pollinations still hasn't given up on providing all kinds of big model services for free, and provides richer models and ways to use them, so I re-recommend it!

Function List

• Image Generation: Enter a text description to generate a unique image.
• Text Generation: Input prompt words to generate natural language text.
• Audio Generation: Supports text-to-speech and speech-to-text.
• API Support: Provides an easy-to-use, free API interface.
• Multiple model options: Includes Flux, Gemini 2.0 Flash, and many more.
• Open source code: users can download the code for self-hosting.
• Automated development: implement new features based on GitHub requests with the MentatBot AI assistant.
• React Integration: Provides React Hooks for developers to use.

Using Help

Pollinations is designed to make it easy for anyone to get started. Here's a detailed how-to, covering the web version, API calls, and self-hosted processes.

Use of the web version

1. Open your browser and visit https://pollinations.aiThe
2. Enter a description in the input box, such as "Forest at Sunset" to generate a picture, or "Write a Short Essay" to generate a text.
3. Click on "Generate" and the result will be displayed after a few seconds.
4. Images can be downloaded and text can be copied.

For audio generation:

1. interviews https://text.pollinations.aiThe
2. Enter text such as "Welcome to Pollinations" and select model=openai-audio, click Generate.
3. Download the generated voice file.

The web version is easy to use and suitable for novice experience.

Mobile Client Version

Android: https://github.com/g-aggarwal/Pollinator/releases/tag/beta-1

Apple: https://apps.apple.com/us/app/pal-chat-ai-chat-client/id6447545085?platform=iphone

API Calls

Image Generation

1. Enter it in your browser:https://pollinations.ai/p/你的描述The
   • Example:https://pollinations.ai/p/蓝色海洋The
2. Parameter adjustment can be added:
   • &width=800&height=600 Set the size.
   • &seed=42 Fixed results.
3. Returns the image URL for direct use.

Text Generation

1. Use the URL:https://text.pollinations.ai/你的提示词The
   • Example:https://text.pollinations.ai/什么是AIThe
2. Returns the text result.

Audio Generation

1. Use the URL:https://text.pollinations.ai/文字?model=openai-audio&voice=novaThe
   • Example:https://text.pollinations.ai/你好?model=openai-audioThe
2. Returns the audio file.

Python Example:

import requests
url = "https://pollinations.ai/p/夜晚星空"
response = requests.get(url)
with open("image.jpg", "wb") as f:
f.write(response.content)
print("图片已下载！")

React Hooks Integration

1. Install the library:

   npm install @pollinations/react
   

2. Sample code:

   import { usePollinationsImage } from '@pollinations/react';
   const MyComponent = () => {
   const imageUrl = usePollinationsImage("巴黎夜景", { width: 800 });
   return <img src={imageUrl} alt="生成图片" />;
   };
   

Pure API Configuration

This is where we officially enter the welfare phase.

Full API documentation: https://github.com/pollinations/pollinations/blob/master/APIDOCS.md

## Pollinations.AI Cheatsheet for Coding Assistants

### Image Generation
Generate Image: `GET https://image.pollinations.ai/prompt/{prompt}`

### Image Models
List Models: `GET https://image.pollinations.ai/models`

### Text Generation
Generate (GET): `GET https://text.pollinations.ai/{prompt}`

### Text Generation (Advanced)
Generate (POST): `POST https://text.pollinations.ai/`

### Audio Generation
Generate Audio: `GET https://text.pollinations.ai/{prompt}?model=openai-audio&voice={voice}`

### OpenAI Compatible Endpoint
OpenAI Compatible: `POST https://text.pollinations.ai/openai`

### Text Models
List Models: `GET https://text.pollinations.ai/models`

### Real-time Feeds
Image Feed: `GET https://image.pollinations.ai/feed`
Text Feed: `GET https://text.pollinations.ai/feed`
*\* required parameter*

for example, in Page Assist Configure the API without filling in the key

Page Assist There is a problem with the automatic model name acquisition in AI Dialog, it does not list the real model list and some models are not directly usable after selection. If you switch to other AI dialog tools and configure the model names manually, you won't have this problem. For the list of configurable model names, please refer to: https://text.pollinations.ai/models

Some models are not available after configuration, may be the official amount is not enough, or the model name is configured incorrectly, sometimes the model name has "-", to be changed to a space.

Self-hosted installation

1. Make sure Git and Python are installed.
2. Cloning code:

   git clone https://github.com/pollinations/pollinations.git
   

3. Go to the catalog:

   cd pollinations
   

4. Install the dependencies:

   pip install -r requirements.txt
   

5. Start the service:

   python main.py
   

6. interviews http://localhost:5000The

Self-hosting is suitable for users who need local operation or higher privacy.

Featured Function Operation

• Model Selection: The API is used in the model=flux Specifies the image model, or model=gemini Specifies the text model.
• audio conversion: Use of openai-audio Model to generate speech, or upload audio to text.
• MentatBot Development: in https://github.com/pollinations/pollinations/issues Submit a feature request and the AI assistant will automatically fulfill it.
• Pollinations.DIY: Access https://pollinations.diy, develop AI apps in the browser.

These features make Pollinations flexible and powerful.

application scenario

1. art
   • The artist typed in "Abstract City" to generate the inspiration image.
2. Educational support
   • Teachers create course materials using text generation tools.
3. application development
   • Developers embed APIs into applications to generate dynamic content.
4. Entertainment Interaction
   • Users generate "science fiction stories" or "cat voices" to entertain themselves.

QA

1. Does Pollinations charge a fee?
   • No charge, totally free, no registration required.
2. Is the generated content commercially available?
   • Yes, but you need to confirm the license of the model used.
3. How do you ensure consistent results?
   • Add &seed=固定值The
4. What languages are supported?
   • Supports multi-language input, output depends on modeling capabilities.
5. How do I contact the team?
   • Submit an issue on GitHub or join the Discord community.

Pollinations Backup of old website information

Pollinations is an Artificial Intelligence image art creation platform that provides the latest AI models and a simple interface that allows users to easily create and explore various forms of images and videos. Users can generate and customize their own content using text prompts, model selection, parameter adjustments, and more, or use the image feed to view other users' work.Pollinations also supports integration with ChatGPT for conversation-based image generation and interaction.Pollinations is not only a creative tool, but also a collaborative ecosystem where users Pollinations is not only a creative tool, but also a collaborative ecosystem where users can discuss, get help, and contribute to the Discord community.Pollinations' goal is to empower many forms of human expression and explore the possibilities and potential of artificial intelligence.

Function List

• Generate images and videos using text prompts
• Selection of different AI models and parameters
• View and share image feeds
• Conversational image generation and interaction in conjunction with ChatGPT
• Join the Discord community to collaborate and learn

Using Help

• Visit [official website] for more information and tutorials
• In [image sourceEnter the topics or keywords you are interested in to view other users' works!
• In [Create pageEnter the text prompts of the image or video you want to generate in [Results]，select the appropriate model and parameters, click the Generate button, and wait for the results to be displayed.
• Enter the character or theme you want to talk to in [ChatGPT page], select the appropriate model and parameters, and click the start button to interact with the AI
• In [Discord CommunityJoin us in the discussion in ] to get help and feedback and share your creations and ideas!

You don't need to check out the above help on how to use it, the official website only offers the following at the moment:

https://pollinations.ai/p/{prompt}

{prompt} Instead of the English image prompt word, you can get the corresponding image URL

Have you figured out the actual use of the service by this point? I'll give a direct example below:

The user will give you a prompt and you must only return your response in that format (don't try to encode anything).

Generating your image...
! [](https://pollinations.ai/p/{prompt}

Important: never respond with anything more than the above text - make sure to stop your generation as soon as you complete the image url.

Pollinations Generation Example

If your interface supports loading standard Makrdown format code and the large language model used is understood correctly, you will get an image. The resulting image is fixed in size and contains a copyrighted logo in the lower right corner, which you can remove through other services provided by Pollinations or by using a dedicated logo removal tool.