perplexity-cli v1.0.0

perplexity-cli

A command-line interface (CLI) app that allows you to run queries using Perplexity's API directly from your terminal.

Installation

Local Installation

# Clone the repository
git clone https://github.com/yourusername/perplexity-cli.git
cd perplexity-cli

# Install dependencies
npm install

# Link the CLI globally
npm link

Global Installation (once published)

npm install -g perplexity-cli

Or run directly without installation:

npx perplexity-cli

Setup

Before using the CLI, you need to set your Perplexity API key:

perplexity-cli set-key YOUR_API_KEY_HERE

You can get your API key from the Perplexity API Settings page.

Usage

Running Queries

perplexity-cli query "What is the capital of France?"

Stream the Response in Real-Time

perplexity-cli query "Explain quantum computing" --stream

Save the Response to a File

perplexity-cli query "Write a short story about AI" --output story.txt

Specify a Different Model

perplexity-cli query "Explain quantum computing" --model sonar-large

View Query History

perplexity-cli history

View Available Models

perplexity-cli models

View Your API Key (Masked)

perplexity-cli view-key

Clear Your API Key

perplexity-cli clear-key

Available Commands

• set-key <key>: Set your Perplexity API key
• view-key: View your currently set API key (masked)
• query <question>: Send a query to the Perplexity API
• models: List available Perplexity API models
• history: View history of recent queries
• clear-key: Clear your stored API key

Options for Query Command

• -m, --model <model>: Specify the model to use for your query (default: sonar)
• -s, --stream: Stream the response as it is generated
• -o, --output <file>: Save the response to a file
• -h, --help: Display help information
• -V, --version: Display version information

Models

The CLI supports various Perplexity models:

• sonar: Default, balanced speed and capability
• sonar-small: Fastest, least capable
• sonar-medium: Good balance of speed and capability
• sonar-large: Most capable, slower
• codellama-70b: Specialized for code generation
• mistral-7b: Open-source model
• mixtral-8x7b: Mixture of experts model
• llama-3-70b: Meta's latest model

License

MIT