AI CLI

AI assistant in your terminal.

$ ai "Tell me an interesting fact about the Roman Empire"
me: Tell me an interesting fact about the Roman Empire
ai: The Roman Empire had a vast network of roads extending over 250,000 miles at its peak...
me: _

Installation

npm install -g @callstack/ai-cli

Setup

API key

You will need a valid API key from one of supported AI inference providers: OpenAI or Perplexity. Using these APIs is paid but the costs should be very low, typically much less than $0.01 per interaction.

Quick config

Run ai init to create a basic ~/.airc.json config file with your OpenAI or Perplexity API keys.

$ ai init

Read more about the configuration file format and available option.

Usage

Run ai [query] to start a chat with an AI assistant. You can put the question in the params.

$ ai "Tell me a joke"
me: Tell me a joke
ai: Why don’t skeletons fight each other? They don’t have the guts.

CLI options

CLI options are passed when invoking the ai commend:

• --help: display available options
• --version: display CLI version
• --provider [name] (or -p [name]): select an inference provider to use:
  • openai
  • anthropic (or anth)
  • perplexity (or pplx)
  • mistral
• --model [name] (or -m [name]): select a model to use. This should be a model available for the selected provider.
• --creative: respond in a creative way
• --precise: respond in a more accurate way
• --verbose (or -V): enable verbose output

You should have a relevant API key in your ~/.airc.json file.

Models

Open AI

Model	Alias	Price: in | out *	Notes
gpt-4o		$5 | $15	Default
gpt-4-turbo		$10 | $30
gpt-4		$30 | $60
gpt-3.5-turbo	gpt-3.5	$0.5 | $1.5

* API prices per 1 million input/output tokens

More info: OpenAI docs

Anthropic

Model	Alias	Price: in | out *	Notes
claude-3-opus-20240229	opus	$15 | $75
claude-3-sonnet-20240229	sonnet	$3 | $15	Default
claude-3-haiku-20240307	haiku	$0.25 | $1.25

* API prices per 1 million input/output tokens

More info: Anthropic docs

Perplexity

Model	Alias	Price: in | out *	Notes
llama-3-sonar-large-32k-chat	large	$1 | $1	Default
llama-3-sonar-large-32k-online	online	$1 | $1 | request: $5	Online**
llama-3-sonar-small-32k-chat	small	$0.2 | $0.2
llama-3-sonar-small-32k-online		$0.2 | $0.2 | request: $5	Online**
llama-3-70b-instruct	llama-3	$1 | $1
llama-3-8b-instruct		$0.2 | $0.2
mixtral-8x7b-instruct	mixtral	$0.6 | $0.6

* API prices per 1 million input/output tokens, per 1 thousands requests

** Perplexity online models can lookup information on the Internet, they also charge a per request fee ($5/1000 requests) and ignore system prompts.

More info: Perplexity docs

Mistral

Model	Alias	Price: in | out *	Notes
mistral-large-latest	large	$4 | $12	Default
mistral-medium-latest	medium	$2.7 | $8.1
mistral-small-latest	small	$1 | $3
open-mixtral-8x22b	mixtral	$2 | $6
open-mixtral-8x7b		$0.7 | $0.7
open-mistral-7b	mistral	$0.25 | $0.25

* API prices per 1 million input/output tokens

More info: Mistral docs

CLI commands

CLI commands are available when using CLI in interactive mode.

• /help: display available commands
• /exit: exit the CLI
• /info: display information about current session
• /forget: forget previous messages in the session
• /verbose: enable/disable verbose mode

License

MIT