Skip to main content
The Avala CLI is an optional command-line tool that provides terminal access to the Avala API. It’s useful for scripting, automation, and quick interactions with your Avala resources.

​
Installation

Install the CLI by adding the [cli] extra when installing the Avala SDK: 
pip install avala[cli]
This installs the avala command and its dependencies (including Click).

​
Configuration

Before using the CLI, configure your API credentials: 
avala configure
You’ll be prompted for:
  • API Key: Your Avala API key
  • Base URL: The API endpoint (defaults to https://api.avala.ai/api/v1)
The command outputs shell export statements that you can add to your .bashrc, .zshrc, or equivalent: 
export AVALA_API_KEY=your-key-here
export AVALA_BASE_URL=https://api.avala.ai/api/v1

​
Alternative: Command-line flags

You can also pass credentials as flags to any command: 
avala --api-key YOUR_KEY datasets list
avala --base-url https://custom.api.com --api-key YOUR_KEY projects list

​
Environment variables

The CLI reads these environment variables:
  • AVALA_API_KEY: Your API key
  • AVALA_BASE_URL: Custom API endpoint (optional)

​
Available commands

The Avala CLI provides commands for managing all major resources:
  • datasets — Create and manage datasets
  • projects — Manage annotation projects
  • tasks — View tasks
  • exports — Create and download exports
  • agents — Manage agents and view executions
  • storage-configs — Configure cloud storage backends
  • inference-providers — Manage inference providers
  • auto-label — Create and monitor auto-labeling jobs
  • quality-targets — Set and evaluate quality metrics
  • consensus — Compute consensus scores
  • webhooks — Manage webhooks and view deliveries
  • fleet — Fleet management (devices, recordings, events, rules, alerts)
Each command group has subcommands like list, get, create, and delete.

​
Usage

General pattern: 
avala [OPTIONS] COMMAND [ARGS]...

​
Examples

List all datasets: 
avala datasets list
Get details for a specific project: 
avala projects get <project-uid>
Create a new webhook: 
avala webhooks create --target-url https://example.com/hook --events task.completed,task.created

​
Getting help

Use --help on any command or subcommand: 
avala --help
avala datasets --help
avala datasets create --help

​
Output format

The CLI displays results in human-readable tables and detail views. For scripting, consider using the Python SDK directly for programmatic access.

Build docs developers (and LLMs) love

Get started for freeTalk to us