Saptiva Studio documentation

Saptiva Studio is the console where a team builds AI agents and workflows, works in GPU labs and uses models through InferNode. This documentation is organized in four guides, one per type of reader.

Members of a ClientUser guideAgents, workflows, GPU labs, models and InferNode, common errors.Client administratorsAdministrationMembers and roles, SSO, shared credentials, infrastructure dashboard, model catalog, quotas.On-prem operatorsPlatformRequirements, installation with internet access and air-gapped, upgrades, backup, observability, security.IntegratorsReferenceAgents, Workflows, Labs, credentials and InferNode APIs; limits; error codes; glossary.

Where to start

• If it's your first day: Accessing Studio and the Console tour.
• If you want to see something working in 10 minutes: your first agent, your first workflow or your first lab.
• If something fails: Errors and troubleshooting, organized by what you see on screen.
• If you are going to call the models from your code: Use the API from code.

Versions

The documentation follows Studio versions. The version selector is in the top bar; the differences between versions are in Release notes. The screenshots match the version shown on each page.