AIGatewayModelFormat
objectRequest and response format supported by this model.
The format type.
Example:openai
AIGatewayTargetModel
objectA target model instance that this model entry routes requests to.
The name of the model defined in the upstream provider that will be executed.
Example:gpt-5-model
Lightweight object used to reference a provider instance, identified by its ref.
Show Child Parameters
One OfConfiguration for a target model.
Anthropic-specific configuration for a model.
Show Child Parameters
AIGatewayTargetModelConfig
Configuration for a target model.
One OfThe number of dimensions for embedding outputs.
The maximum number of tokens to generate in the response.
Cost per input token for billing and cost tracking.
Cost per output token for billing and cost tracking.
Controls randomness in the model output. Higher values produce more varied responses.
Limits the number of highest-probability tokens considered during generation.
Nucleus sampling probability mass. Tokens with cumulative probability up to top_p are considered.
The upstream URL for the model endpoint.
Allowed values:anthropic
The Anthropic API version to use.
AIGatewayTargetModelAnthropicConfig
objectAnthropic-specific configuration for a model.
The number of dimensions for embedding outputs.
The maximum number of tokens to generate in the response.
Cost per input token for billing and cost tracking.
Cost per output token for billing and cost tracking.
Controls randomness in the model output. Higher values produce more varied responses.
Limits the number of highest-probability tokens considered during generation.
Nucleus sampling probability mass. Tokens with cumulative probability up to top_p are considered.
The upstream URL for the model endpoint.
Allowed values:anthropic
The Anthropic API version to use.
AIGatewayTargetModelAzureConfig
objectAzure-specific configuration for a model.
The number of dimensions for embedding outputs.
The maximum number of tokens to generate in the response.
Cost per input token for billing and cost tracking.
Cost per output token for billing and cost tracking.
Controls randomness in the model output. Higher values produce more varied responses.
Limits the number of highest-probability tokens considered during generation.
Nucleus sampling probability mass. Tokens with cumulative probability up to top_p are considered.
The upstream URL for the model endpoint.
Allowed values:azure
The Azure deployment ID for the model.
The Azure OpenAI API version to use.
Default:2023-05-15