Workspaces are isolated environments for organizing your ML resources. Each workspace has its own models, datasets, workflows, secrets, API keys, and team members.

Quick Start

# View current workspace
mixtrain config

# Switch workspace
mixtrain config --workspace production

Key Concepts

Workspace Isolation

Each workspace contains:

• Models and inference runs
• Datasets
• Workflows
• Secrets and API keys
• Team members with roles

Switching Workspaces

In the Web UI: Click the workspace selector in the top navigation.

Via CLI:

mixtrain config                          # View current
mixtrain config --workspace production   # Switch

Via SDK:

from mixtrain import MixClient

# Use default workspace
client = MixClient()

# Specify workspace explicitly
client = MixClient(workspace_name="production")

Creating Workspaces

mixtrain workspace create staging --description "Staging environment"

Membership and Roles

Manage members in the web UI under Workspace Settings > Members.

API Keys

API keys are workspace-scoped. See Authentication for creating and managing keys.

Secrets

Store encrypted credentials for use in workflows. See Secrets.

mixtrain secret set openai_key -d "OpenAI API key"

Providers

Configure model providers per workspace. See Integrations.

mixtrain provider add anthropic
mixtrain provider status

Next Steps

• Secrets Guide - Managing workspace secrets
• Integrations - Configuring providers