Skip to main content
Manage user access to your GPU clusters by adding team members who can SSH into Slurm clusters or access Kubernetes resources. Learn more about GPU Clusters →

​
User Roles and Permissions

GPU Cluster projects support two user roles with different permission levels:

​
Admin

Admins have full control over both control plane and data plane operations. Control Plane Permissions (Full Write Access):
  • Create clusters
  • Delete clusters
  • Create storage volumes
  • Delete storage volumes
  • Modify cluster configurations
  • Scale up or down nodes
Data Plane Permissions (Full Access):
  • SSH into cluster nodes
  • Run workloads and jobs
  • Access Kubernetes Dashboard
  • Execute kubectl commands
User Management:
  • Add members to the project
  • Remove members from the project
  • Assign user roles

​
Member

Members have read-only access to control plane resources but full access to data plane operations. Control Plane Permissions (Read-Only):
  • ✓ View clusters
  • ✓ View storage volumes
  • ✓ View cluster configurations
  • ✗ Cannot create clusters
  • ✗ Cannot delete clusters
  • ✗ Cannot create storage volumes
  • ✗ Cannot delete storage volumes
  • ✗ Cannot scale the cluster - up or down
Data Plane Permissions (Full Access):
  • SSH into cluster nodes
  • Run workloads and jobs
  • Access Kubernetes Dashboard
  • Execute kubectl commands
  • Deploy pods and manage resources
User Management:
  • ✗ Cannot add or remove users
Key Difference: Members can fully utilize clusters for their work (SSH, run jobs, scale nodes) but cannot create or destroy infrastructure resources.

​
Adding Users

Prerequisites: Users must have a Together AI account and an SSH key added in the Together Playground before they can be added to your cluster.

​
Step-by-Step Instructions

  1. Access Settings
    • Log in to your Together AI account at api.together.ai
    • Click your avatar in the top-right corner
    • Select Settings from the dropdown menu
  2. Navigate to GPU Cluster Projects
    • In the left sidebar, click GPU Cluster Projects
  3. Select Your Project
    • Click View Project on the project you want to manage
  4. Add a New User
    • Click the Add User button
    • A popup dialog will appear
  5. Enter User Email
    • Enter the email address of the user you want to add
    • Click Add User to confirm
  6. Verify Addition
    • If successful, the user will appear in the members grid
    • If the user doesn’t have a Together AI account or SSH key, you’ll see an error message and the user cannot be added

​
Removing Users

To remove a user from your cluster:
  1. Navigate to Settings > GPU Cluster Projects
  2. Click View Project on the relevant project
  3. Find the user in the members grid
  4. Click the three dots (⋯) on the right side of their row
  5. Select Remove User from the dropdown menu
  6. Confirm the removal when prompted