Access the world's best open language models in one place!

OllamaFreeAPI provides free access to leading open-source LLMs including:

• 🦙 LLaMA (Meta)
• 🌪️ Mistral (Mistral AI)
• 🔍 DeepSeek (DeepSeek)
• 🦄 Qwen (Alibaba Cloud)

No payments. No credit cards. Just pure AI power at your fingertips.

# Install or upgrade to the latest version
pip install ollamafreeapi --upgrade

• API Reference - Complete API documentation
• Usage Examples - Practical code examples
• Model Catalog - Available models and their capabilities

Here are some key statistics about the current state of OllamaFreeAPI:

• Active Models: 16 (Ready to use and tested)
• Model Families: 3 (gemma, llama, qwen)
• Endpoints: 6 highly reliable server nodes

from ollamafreeapi import OllamaFreeAPI

client = OllamaFreeAPI()

# Stream responses in real-time
for chunk in client.stream_chat('What is quantum computing?', model='llama3.2:3b'):
    print(chunk, end='', flush=True)

from ollamafreeapi import OllamaFreeAPI

client = OllamaFreeAPI()

# Get instant responses
response = client.chat(
    model="gpt-oss:20b",
    prompt="Explain neural networks like I'm five",
    temperature=0.7
)
print(response)

• llama3.2:3b - Meta's efficient 3.2B parameter model
• deepseek-r1:latest - Strong reasoning capabilities built on Qwen
• gpt-oss:20b - Powerful Gemma-based 20B completion model
• mistral:latest - High-performance baseline Mistral model

• mistral-nemo:custom - 12.2B open weights language model
• bakllava:latest - Vision and language model
• smollm2:135m - Extremely lightweight assistant

Our free API is powered by distributed community nodes:

• Fast response times
• Automatic load balancing and server selection
• Real-time availability checks

# List available models
api.list_models()  

# Get model details
api.get_model_info("mistral:latest")  

# Generate text
api.chat(model="llama3.2:3b", prompt="Your message")

# Stream responses
for chunk in api.stream_chat(prompt="Hello!", model="llama3:latest"):
    print(chunk, end='')

# Check server locations
api.get_model_servers("deepseek-r1:latest")

# Generate raw API request
api.generate_api_request(model="llama3.2:3b", prompt="Hello")

# Get random model parameters (useful for LangChain integration)
api.get_llm_params()

We welcome contributions! Please see our Contributing Guide for details.

Open-source MIT license - View License

• Documentation
• Examples
• GitHub Issues