--- title: "Get an AI Gateway model group" method: GET path: "/api/v2/model-groups/{id}" tags: ["AI Gateway"] --- # Get an AI Gateway model group `GET /api/v2/model-groups/{id}` ## Request - Base URL: `https://api.datadoghq.com` - URL: `https://api.datadoghq.com/api/v2/model-groups/{id}` - Auth: one of: - API key in header `DD-API-KEY` and API key in header `DD-APPLICATION-KEY` - OAuth 2 - Authorization code: `/oauth2/v1/authorize` - Token URL: `/oauth2/v1/token` scopes: `ai_gateway_config_read` - TypeScript: (a function with typed inputs; its response is checked against the schema) Get a named policy or the explicit organization-wide policy by id "default". Requires configuration read permission; named policies also require view access. Provider selections cover current and future models and are returned separately from individual model selections. ## Path parameters - `id` string, required ## Response `200` OK - AiGatewayModelGroupResponse: Response containing one AI Gateway model group. - `data` AiGatewayModelGroupResource, required: AI Gateway model group resource. - `attributes` AiGatewayModelGroupAttributes, required: Complete AI Gateway model group configuration. - `allow_fast_mode` boolean, required: Whether fast mode is allowed for models and routing rules in the group. - `created_at` string, date-time, required: Time when the resource was created. - `max_reasoning_level` 'none' | 'minimal' | 'low' | 'medium' | 'high' | 'xhigh' | 'max', nullable, required: Maximum reasoning level permitted by a model group. - `model_ids` integer[], required: All saved model catalog IDs, including selections whose provider metadata is hidden. Preserve these IDs independently of the visible catalog. - `models` AiGatewayModelGroupModel[], required: Metadata for saved model selections whose providers the caller can view. Missing metadata does not remove a saved ID. - `created_at` string, date-time, required: Time when the resource was created. - `enabled` boolean, required: Whether the resource is enabled. - `id` string, required: Model catalog entry identifier. - `metadata` AiGatewayModelMetadata: Optional, typed enrichment for a model catalog entry. Missing properties remain unknown. - `capabilities` AiGatewayModelCapabilities, required: Model features reported by the enrichment source. Omitted properties are unknown. - `attachment` boolean: Whether attachment input is supported. - `interleaved_reasoning` boolean: Whether reasoning can be interleaved with tool use. - `reasoning` boolean: Whether reasoning is supported. - `structured_output` boolean: Whether structured output is supported. - `temperature` boolean: Whether temperature can be configured. - `tool_call` boolean: Whether tool calls are supported. - `description` string: Human-readable model description. - `display_name` string: Human-readable model name. - `family` string: Model family identifier. - `kind` 'language' | 'embedding' | 'image' | 'audio' | 'decision' | 'reranking': Primary model kind. Decision models return a choice or score instead of generated text. - `knowledge_cutoff` string: Knowledge cutoff reported by the source. - `last_updated` string: Date the source record was last updated. - `lifecycle` 'active' | 'alpha' | 'beta' | 'deprecated': Model lifecycle state. - `limits` AiGatewayModelLimits, required: Token limits reported by the enrichment source. - `context_window_tokens` integer: Total model context window in tokens. - `max_input_tokens` integer: Maximum input tokens when reported separately. - `max_output_tokens` integer: Maximum output tokens. - `modalities` AiGatewayModelModalities, required: Supported model input and output modalities. - `input` string[] - `output` string[] - `open_weights` boolean: Whether model weights are openly available. - `reasoning` AiGatewayModelReasoning: Supported reasoning controls. Absence means controls are unknown or unavailable. - `controls` AiGatewayModelReasoningControl[], required - … - `reference_pricing` AiGatewayModelReferencePrice[]: Provider-specific reference prices from the enrichment source. Omitted for shared canonical-model fallback matches. These do not determine gateway charges. - `amount` string, required: Exact decimal amount per unit, zero or greater, represented as a string to preserve precision. Supports fixed decimal and exponent notation. - `component` string, required: Token category priced, such as input, output, cache_read, or cache_write. - `context_max_tokens` integer: Largest input length covered by this price. Omitted when there is no upper bound. - `context_min_tokens` integer, required: Input length above which this price applies. Zero identifies the base tier. - `currency` string, required: Currency of the price. The only supported value is USD. - `mode` string, required: Source mode name. Standard is the base mode; other values, such as fast, are source-defined. - `service_tier` string: Provider service tier explicitly associated with this mode by the source. Omitted when unknown. - `unit` string, required: Unit of the quoted amount. The only supported value is million_tokens. - `release_date` string: Model release date. - `source` AiGatewayModelMetadataSource, required: Provenance for model metadata enrichment. - `matched_by` 'override' | 'wire_id' | 'canonical_id' | 'wire_id_date_suffix' | 'canonical_model', required: Identifier strategy used to match the gateway catalog entry. Date-suffix and shared canonical-model fallback matches produce partial metadata. - `model_id` string, required: Model identifier in the metadata source provider catalog. - `name` string, required: Metadata source name. - `observed_at` string, date-time: Last successful observation of this valid source record. Retained invalid entries keep their previous timestamp. - `provider` string, required: Provider namespace in the metadata source, or canonical for a shared model record. - `revision` string: SHA-256 content revision of the source snapshot that supplied this record. - `status` 'available' | 'partial', required: Completeness of the matched source record. Partial also identifies metadata from a base model for a dated snapshot or from a shared canonical-model fallback. - `model_id_canonical` string, required: Provider-independent model identifier. - `model_id_wire` string, required: Model identifier sent to the provider. - `provider_id` integer, required: Provider identifier. - `source` 'discovered' | 'manual', required: How a model catalog entry is managed. - `supported_operations` AiGatewaySupportedOperation[], required: Operations supported by the model. - `updated_at` string, date-time, required: Time when the resource was last updated. - `name` string, required: Human-readable resource name. - `provider_ids` integer[], required: All explicitly selected provider IDs, including hidden providers. Embedded provider metadata includes only viewable providers. - `providers` AiGatewayModelGroupProvider[], required: Viewable explicitly selected providers, separate from individually selected models. Use provider_ids to retain all selections. - `id` integer, required: Provider ID. - `name` string, required: Provider display name. - `revision` integer: Revision of the organization-wide policy. Present for the default model group and absent for named policies. - `routing_rules` AiGatewayModelGroupRoutingRule[], required - `slug` string, required: Organization-unique routing rule slug. - `updated_at` string, date-time, required: Time when the resource was last updated. - `id` string, required: Resource identifier. - `type` 'model-groups', required: AI Gateway model group resource type. ## Other responses - `400`: Bad Request - `403`: Forbidden - `404`: Not Found - `429`: Too many requests --- [API](https://skmtc.dev/datadog/apis/datadog-api-v2-collection.md) - [All operations](https://skmtc.dev/datadog/apis/datadog-api-v2-collection/llms.txt) - [OpenAPI document](https://skmtc.dev/datadog/apis/datadog-api-v2-collection/revisions/3f70db624329?raw)