GoodMemGoodMem
ReferenceSdkRustTypes

LlmCreationRequest

Rust struct goodmem::models::LlmCreationRequest: declaration, fields or variants, and implementations.

All types

Struct LlmCreationRequest

goodmem::models::LlmCreationRequest

Request body for creating a new LLM. An LLM represents a configuration for text generation services.

pub struct LlmCreationRequest {
    pub display_name: String,
    pub description: Option<String>,
    pub provider_type: LlmProviderType,
    pub endpoint_url: String,
    pub api_path: Option<String>,
    pub model_identifier: String,
    pub supported_modalities: Option<Vec<Modality>>,
    pub credentials: Option<Box<EndpointAuthentication>>,
    pub labels: Option<BTreeMap<String, String>>,
    pub version: Option<String>,
    pub monitoring_endpoint: Option<String>,
    pub capabilities: Option<Box<LlmCapabilities>>,
    pub default_sampling_params: Option<Box<LlmSamplingParams>>,
    pub max_context_length: Option<i32>,
    pub client_config: Option<BTreeMap<String, serde_json::Value>>,
    pub owner_id: Option<String>,
    pub llm_id: Option<String>,
    pub dashscope_api_dialect: Option<DashScopeApiDialect>,
}

Referenced types: DashScopeApiDialect, EndpointAuthentication, LlmCapabilities, LlmProviderType, LlmSamplingParams, Modality.

Fields

display_name

pub display_name: String

User-facing name of the LLM Wire name: displayName.

description

pub description: Option<String>

Description of the LLM Wire name: description.

provider_type

pub provider_type: LlmProviderType

Type of LLM provider Wire name: providerType.

endpoint_url

pub endpoint_url: String

API endpoint base URL (OpenAI-compatible base, typically ends with /v1) Wire name: endpointUrl.

api_path

pub api_path: Option<String>

API path for chat/completions request (defaults to /chat/completions if not provided) Wire name: apiPath.

model_identifier

pub model_identifier: String

Model identifier Wire name: modelIdentifier.

supported_modalities

pub supported_modalities: Option<Vec<Modality>>

Supported content modalities (defaults to TEXT if not provided) Wire name: supportedModalities.

credentials

pub credentials: Option<Box<EndpointAuthentication>>

Structured credential payload describing how to authenticate with the provider. Required for SaaS providers; optional for local or proxy providers. Wire name: credentials.

labels

pub labels: Option<BTreeMap<String, String>>

User-defined labels for categorization Wire name: labels.

version

pub version: Option<String>

Version information Wire name: version.

monitoring_endpoint

pub monitoring_endpoint: Option<String>

Monitoring endpoint URL Wire name: monitoringEndpoint.

capabilities

pub capabilities: Option<Box<LlmCapabilities>>

LLM capabilities defining supported features and modes. Optional - server infers capabilities from model identifier if not provided. Wire name: capabilities.

default_sampling_params

pub default_sampling_params: Option<Box<LlmSamplingParams>>

Default sampling parameters for generation requests Wire name: defaultSamplingParams.

max_context_length

pub max_context_length: Option<i32>

Maximum context window size in tokens Wire name: maxContextLength.

client_config

pub client_config: Option<BTreeMap<String, serde_json::Value>>

Provider-specific client configuration as flexible JSON structure Wire name: clientConfig.

owner_id

pub owner_id: Option<String>

Optional owner principal UUID. If omitted, defaults to the authenticated principal. CREATE_LLM is evaluated against the proposed LLM and owner. Wire name: ownerId.

llm_id

pub llm_id: Option<String>

Optional client-provided UUID for idempotent creation. If not provided, server generates a new UUID. Returns ALREADY_EXISTS if ID is already in use. Wire name: llmId.

dashscope_api_dialect

pub dashscope_api_dialect: Option<DashScopeApiDialect>

DashScope request and response API dialect. Valid only for the DASHSCOPE provider. When omitted, a recognized apiPath determines the dialect; otherwise OPENAI_COMPATIBLE is used. Wire name: dashscopeApiDialect.

Implementations

validate

pub fn validate(&self) -> crate::Result<()>

Validate the request before sending, without making an HTTP call.

from_model

pub fn from_model(model: &str, api_key: Option<&str>) -> Result<Self>

Fill known provider defaults; explicit request structs support custom models.

Trait implementations

  • Clone (derived)
  • Default (derived)
  • PartialEq (derived)
  • Serialize (derived)
  • impl<'de> Deserialize<'de> for LlmCreationRequest
  • impl std::fmt::Debug for LlmCreationRequest

On this page