LLM CLI

LLM-CLI is a command-line AI assistant that's a fork of Gemini Cli. Unlike its predecessor, LLM-CLI is a multi-provider tool that supports Gemini, OpenAI, Anthropic Claude, Ollama, and more, making it a powerful agent for coding. This means you can now use all of Gemini Cli’s powerful features with your preferred large language model—whether it's running in the cloud or locally.

Installation

npm install -g @marizmelo/llm-cli

Usage

llm-cli

For non-interactive mode:

llm-cli --prompt "What is 2+2?"

Features

• 🤖 Multi-provider support (OpenAI, Anthropic, Ollama, Gemini)
• 🔧 Interactive CLI interface with full tool/function support
• 📁 Context-aware file processing
• 🔌 Extensible with custom commands
• 🎯 Provider-specific model switching
• 💾 Memory and context management
• 🔑 Secure API key management with persistent settings
• 🛠️ Full tool/function calling support for OpenAI, Anthropic, and Ollama

Provider Setup

Quick Setup

# Setup cloud providers with persistent API keys
llm-cli
/provider setup openai sk-your-api-key
/provider setup anthropic sk-ant-your-api-key
/provider switch openai

Ollama (Local)

# 1. Install Ollama (https://ollama.com)
# 2. Pull a tool-capable model
ollama pull qwen2.5:latest

# 3. Start llm-cli and switch to Ollama
llm-cli
/provider switch ollama

# 4. (Optional) Switch to a different model
/provider model <model-name>

No .env file or environment variables required for Ollama. The default model is llama3.2:latest. Use /provider model to list available models and switch between them.

Tool-capable Ollama models (recommended for file ops, shell commands, etc.):

• qwen2.5:7b / qwen2.5:14b
• llama3.1:8b / llama3.1:70b
• mistral:7b

Models without tool support (e.g. gemma3n, deepseek-r1) will automatically fall back to text-only conversation mode.

Environment Variables (Alternative)

# Cloud providers
export OPENAI_API_KEY="your-api-key"
export ANTHROPIC_API_KEY="your-api-key"

# Ollama (optional overrides)
export OLLAMA_MODEL="qwen2.5:latest"
export OLLAMA_BASE_URL="http://localhost:11434"

llm-cli

Documentation

For full documentation, visit: https://github.com/marizmelo/llm-cli

License

Apache-2.0