Models

Model Catalog (/workspace/model-catalog)

Overview of providers, keys and usage.

Model Providers (/workspace/providers)

Configure providers and credentials.

• Add a known provider or a custom provider; delete a provider with confirmation.
• Configuration areas per provider: API keys, deployments, network, beta headers.
• A user holding only the Settings permission is redirected to Pricing Overrides.

Budgets & Limits (/workspace/model-limits)

Budgets and rate limits. This screen blocks: a request over the limit is refused. The soft cut (Soft budgets, under the Costs group) is the opposite and never denies.

• Limit creation and edit sheet: provider (or all), model name (or a wildcard), scope, usage limits with token limits and request limits, and current consumption.

Routing Rules (/workspace/routing-rules)

Intelligent routing rules, with an edit sheet per rule.

• A read-only decision tree visualization of the rules exists at /workspace/routing-rules/tree, without its own menu item.

Complexity Router (/workspace/complexity-router)

Routes by complexity band. The screen has a View documentation button that opens this page.

• Configurable: the three band boundaries, shown as SIMPLE to MEDIUM, MEDIUM to COMPLEX and COMPLEX to REASONING, over a complexity spectrum. The band names are wire values and are not translated; labels, descriptions and validation messages are.
• Keyword lists: converted to lowercase and deduplicated on save; each list requires at least one entry.
• Actions: discard changes and restore defaults.

Pricing Overrides (/workspace/custom-pricing/overrides)

Scoped pricing overrides, edited via a dedicated sheet.

Model Settings (/workspace/custom-pricing)

Model and routing configuration.

• Form fields: pricing_datasheet_url, pricing_sync_interval_hours, model_parameters_url, routing_chain_max_depth, live_models_sync_interval_minutes (0 turns off the background refresh).

About Atlasberg Platform

Atlasberg Platform is the control layer between a company and every AI model: each request is authenticated with a virtual key, filtered by policy and DLP, routed to the right provider and written to a hash-chained audit trail. It exposes an OpenAI-compatible API, so applications only swap the base URL.

The same artifact runs in Atlasberg Cloud, in your VPC, on-premises or fully air-gapped, and is priced by capacity and modules, never per seat.

This page is part of the official documentation. To talk to the engineering team, write to [email protected] or use https://atlasberg.com/contato. Answers come within one business day.