OpenRouter

Connect hi-shell to OpenRouter for access to dozens of LLM models through a single API key. OpenRouter provides a unified interface to models from OpenAI, Anthropic, Google, Meta, and more.

Prerequisites

An OpenRouter account
An API key from openrouter.ai/keys

Setup

1. Get an API Key

Sign up at openrouter.ai
Go to Keys
Create a new API key
Copy the key (starts with sk-or-)

2. Configure hi-shell

hi-shell --init
# Select "Cloud" → "OpenRouter"
# Paste your API key
# Choose a model (e.g., "microsoft/phi-3-mini-4k-instruct")

Manual Configuration

Edit ~/.config/hi-shell/config.toml:

llm_provider = "Cloud"
cloud_provider = "OpenRouter"
cloud_model = "microsoft/phi-3-mini-4k-instruct"
api_key = "sk-or-your-api-key-here"

Recommended Models

Model	Cost	Quality	Speed	Best For
`microsoft/phi-3-mini-4k-instruct`	Free	Good	Fast	Default, good balance
`meta-llama/llama-3.2-3b-instruct`	Free	Good	Fast	General use
`google/gemini-2.0-flash-exp`	Free	Good	Very fast	Quick commands
`anthropic/claude-3.5-sonnet`	Paid	Excellent	Medium	Complex tasks
`openai/gpt-4o`	Paid	Excellent	Medium	High quality

Many models on OpenRouter are free to use. Check openrouter.ai/models for pricing.

Configuration Options

Option	Description
`cloud_model`	Model identifier (e.g., `microsoft/phi-3-mini-4k-instruct`)
`api_key`	Your OpenRouter API key

Security Note

Your API key is stored locally in the config file. Keep it secure:

# Set restrictive permissions on your config
chmod 600 ~/.config/hi-shell/config.toml

Troubleshooting

“Invalid API key” Error

Verify your key at openrouter.ai/keys
Make sure the key starts with sk-or-
Check for extra whitespace when pasting

Rate Limiting

Free models have rate limits. If you hit them:

Wait a moment and try again
Switch to a paid model for higher limits
Use a local provider instead

Model Not Available

Some models may be temporarily unavailable. Try:

A different model
Checking the model status at openrouter.ai