Register & Login
End users complete email registration; Operator, End User, and Provider log in and enter their workspaces.
Scenario Navigator
Scenario guides are organized around outcomes: what you want to accomplish. Choose the closest task or scenario, follow the reading order on its overview page, and open the linked user-manual pages when you need page-level field and button details.
| Role | Typical responsibilities |
|---|---|
| Operator | Configure platform capabilities, onboard resources, assign access, process approvals, and review platform monitoring |
| Provider | Publish and maintain models, configure billing and rate limits, and review customer usage and model revenue |
| End User | Discover and try models, request quota, call APIs, and create development, training, or deployment workloads |
If you are unsure which role applies, start with the Role Comparison, then use the role and subsystem filters below. Use each scenario's completion checklist to confirm that the task is actually finished.
Common Tasks
Choose the task closest to your problem to see an ordered reading path. If no task matches, continue with the scenario filters below.
Use Scenarios
Showing 29 scenarios. If no task matches, filter by user role and subsystem.
01
Identity Authorization
Operators control who can see and operate features through roles, menus, and button permissions.
Publish Models
Providers publish external endpoints, self-owned models, or platform-hosted models as public or private services with billing and throttling.
Publish Models Preconfiguration
Prepare standardized meta-models, model sources, templates, and tags to reduce repeated provider configuration.
Publish Aggregation Models
Wrap multiple published models behind one entry point and route requests by success rate, cost, balance, random, or round-robin policies.
Model Experience & API Calling
End users discover, try, integrate, and review calls; providers monitor customer-level calling activity.
Model Usage & Earnings
Convert tokens, duration, and request counts into end-user usage and provider revenue across billing periods.
Recharge & Billing
Obtain spendable balance, record model and card-hour usage, and generate usage, revenue, and billing-period data.
On-Prem Compute Onboarding
Connect private IDC, local GPU, NPU, or XPU resources to AGIOne as schedulable, metered, and observable resource pools.
On-Prem Inference Template Building
Turn inference parameters into templates so users can quickly deploy online inference services.
On-Prem Model Deployment & Status Check
Deploy online inference services on local resource pools and confirm they are runnable, accessible, and diagnosable.
On-Prem Development, Training & Assets
Support development, training, and data management while accumulating models, images, datasets, and other assets.
On-Prem Resource Metering & Monitoring
Control quotas and monitor resource pool runtime status, capacity, usage, and billing-period metering.
On Cloud Resource Access
Connect cloud providers, cloud accounts, regions, and resource pools, then authorize tenants or business types to use them.
On Cloud Model Asset Publishing
Configure runtime environments, frameworks, categories, model information, and output APIs as deployable cloud model assets.
On Cloud Model Deployment & Calling
Select models from the cloud model marketplace, deploy them, and obtain API calling capability.
Model Publishing Approval
Review provider publishing requests for model information, agreements, billing, and throttling governance.
Observability & Troubleshooting
Locate throughput, latency, status, failure, and resource issues from model, organization, and resource perspectives.
Platform Governance & Access Control
Control who can access what, how much they can call, and which keys, projects, models, and quotas they can use.
Billing-Cycle Reconciliation & Settlement
Reconcile monthly close, financial accounts, and exceptions before generating and reviewing settlement statements.
Provider Revenue & Settlement
Explain provider revenue by customer, model, and cycle and reconcile expected settlement, receipt, and transfer.
Project, Key & Budget Governance
Control calling scope, cost, and credential lifecycle with project budgets, model allowlists, and purpose-specific keys.
Member Quota Request & Allocation
Request, adjust, and limit member quota while coordinating member, project, key, and model-allowlist boundaries.
Application Publishing & Approval
Review bound models, calling entries, and customer scope, then validate approval, visibility, and calls.
On-Prem Runtime Images & Storage
Connect image services, runtime images, and block, file, or object storage and validate them with a test workload.
API Rate-Control Release & Audit
Design and release rate-control rules from a traffic baseline and verify node versions, hits, and audit details.
On-Cloud Scheduling Policies
Configure preferred pools, capacity thresholds, and failure fallback and validate actual scheduling with a test deployment.
License Lifecycle Management
Review activation state, validity, authorization composition, and managed objects and prepare renewal or expansion.
On-Prem New Model Template Preparation & Deployment
Complete ModelScope model download, Ascend 910B inference template creation, and validate instance deployment, publishing, and API calls.