FreeInference

Free LLM inference for coding agents and AI-powered IDEs.

Overview

FreeInference provides free access to state-of-the-art language models specifically designed for coding agents like Cursor, Codex, Roo Code, and other AI-powered development tools.

Documentation

Visit our documentation at: https://harvardsys.github.io/free_inference/

Supported IDEs & Coding Agents

• Cursor - AI-powered code editor
• Codex - Terminal-based coding assistant
• Roo Code - VS Code & JetBrains extension
• Kilo Code - AI coding assistant
• And any tool that supports OpenAI-compatible APIs

Quick Start

Cursor Setup

1. Open Settings (Cmd + , or Ctrl + ,)
2. Go to API Keys section
3. Enter your FreeInference API key
4. Click Override OpenAI Base URL
5. Enter: https://freeinference.org/v1
6. Enable the toggle and start coding!

Codex Setup

1. Create ~/.codex/config.toml:

model = "glm-4.7"
model_provider = "free_inference"

[model_providers.free_inference]
name = "FreeInference"
base_url = "https://freeinference.org/v1"
wire_api = "chat"
env_http_headers = { "X-Session-ID" = "CODEX_SESSION_ID", "Authorization" = "FREEINFERENCE_API_KEY" }

2. Add to ~/.zshrc or ~/.bashrc:

export CODEX_SESSION_ID="$(date +%Y%m%d-%H%M%S)-$(uuidgen)"
export FREEINFERENCE_API_KEY="Bearer your-api-key-here"

3. Reload: source ~/.zshrc

Roo Code / Kilo Code Setup

1. Install the extension in your IDE
2. Open settings
3. Select OpenAI Compatible as provider
4. Configure:
   • Base URL: https://freeinference.org/v1
   • API Key: your-api-key-here
5. Select your preferred model

Available Models

• GLM-4.7 - 200K context, best for long context and bilingual support
• GLM-4.7-Flash - 200K context, fast and cost-effective
• MiniMax M2 - 196K context, best for very large codebases
• Qwen3 Coder 30B - 32K context, specialized for code generation
• Llama 3.3 70B - 131K context, general coding (limited capacity)
• Llama 4 Scout - 128K context, optimized for speed (limited capacity)
• Llama 4 Maverick - 128K context, multimodal support (limited capacity)

See the Models documentation for the complete list.

Get API Key

1. Visit https://freeinference.org
2. Register for a free account
3. Log in and create your API key
4. Start using FreeInference with your favorite IDE!

Documentation Links

• Quick Start Guide
• IDE Integration Guides
• Configuration Examples
• Available Models

Support

• Documentation: https://harvardsys.github.io/free_inference/
• Issues: GitHub Issues
• Questions: Contact the team