Instance Catalog
Compare GPU type, memory, CPU/RAM profile, and storage options before launching.
Product / Instances
Launch GPU sessions in minutes.
Self-serve instances for exploration, model prototyping, and burst training runs with real-time queue visibility.
Why teams choose this path
Capabilities built for high-throughput ML operations
- Self-serve notebooks, SSH sessions, and batch launch workflows
- Live queue visibility with clear provisioning and runtime states
- Per-project usage tracking for cost and performance accountability
What you can do
Operational workflows supported on day one
Session Lifecycle
Track workloads from queued to running to completed with clear state transitions and status context.
Usage Controls
Set runtime expectations and monitor spend by project to keep experimentation within budget.