Ways to use distil labs

Four things reach the distil labs platform. Three of them drive it: the CLI, the Claude skill and the REST API. They create the same entities, accept the same config overrides, and identify each entity by the same UUID, so a project can move between them freely. The fourth, the dashboard, only reads.

Way	What you can do	Best for
CLI	Everything: prepare data, run every job, read metrics, deploy	The default. Everything on this site assumes it
Claude skill	In Claude Code, everything the CLI does. Elsewhere, choosing the task type and writing the data files	Your first model, and any task where you want a second opinion
REST API	Everything the CLI does, over HTTP	Windows without WSL, locked-down machines, and Python pipelines
Dashboard	Read your models, datasets and traces, with their statuses and metrics	Watching a run without a terminal

The CLI

The distil CLI is how you drive the platform. Everything on this site assumes it is installed unless a page says otherwise.

curl -fsSL https://cli-assets.distillabs.ai/install.sh | sh

The script writes one binary to ~/.local/bin/distil. It changes nothing else, and it edits no shell configuration file.

Check that it answers:

distil --version

If your shell cannot find the binary, add ~/.local/bin to your PATH and try again.

Then create your account with distil signup: Account and authentication.

Keep it current

The platform moves quickly and the CLI moves with it. Update in place:

distil update

Check your version with distil --version. The CLI reference names the version these docs describe.

Supported platforms

Platform	Supported
Linux (x86_64)	Yes
Linux (arm64)	Yes
macOS (Intel)	Yes
macOS (Apple Silicon)	Yes
Windows	No

There is no Windows binary, and some environments permit no new executables at all. Use the REST API instead. On Windows you can also run the CLI inside WSL.

Command-line JSON

Several examples in these docs pipe --output json through jq. You do not need it to use the platform, but status checks and metrics are much easier to read with it.

brew install jq        # macOS
apt-get install jq     # Debian / Ubuntu

Every command, flag and output shape: CLI reference.

The Claude skill

The distil labs Claude skill teaches Claude to drive this workflow with you: choosing a task type, preparing data, running the pipeline, and analysing what comes back.

In Claude Code:

/plugin marketplace add https://github.com/distil-labs/distil-cli-skill
/plugin install distil-cli@distil-cli-skill

In Claude.ai or Claude Desktop:

Download the skill as a ZIP
Go to claude.ai → Settings → Capabilities → Skills
Click “Upload skill” and select the ZIP file
Toggle the skill ON

Environment	What Claude can do
Claude Code	The full workflow: choose a task type, prepare data, run commands, train, analyse, deploy
Claude.ai	Preparation and analysis: choose a task type and write your data files. You run the commands yourself.

The skill needs the CLI and an account, the same as a run you drive by hand. To train your first model with Claude before you install anything, use the onboarding script from Get started instead.

The REST API

Everything the CLI does, the REST API does. Use it when the CLI cannot be installed: Windows without WSL, an environment that permits no new binaries, or when your work is already scripted in Python.

The API is https://api.distillabs.ai. It authenticates with the same email and password as the CLI, through a Cognito access token that lasts one hour. See Authenticate with the REST API.

Using the REST API - the working guide, with the helpers every example depends on
api.distillabs.ai/docs - the complete endpoint schema

The dashboard

app.distillabs.ai lists your models, seed datasets, training datasets and traces. Each row opens a detail page with that entity’s status, its metrics, and the CLI command that works with it.

The dashboard is read-only. You start every job from the CLI or the REST API. To read your credit balances, run distil credits-balance.

Create your account, then build a model.