Struct LlmConfig 

pub struct LlmConfig {

    pub provider: String,
    pub model: String,
    pub api_key_env: String,
    pub base_url: Option<String>,
    pub max_tokens: usize,
    pub temperature: f32,
    pub context_window: usize,
    pub input_cost_per_million: f64,
    pub output_cost_per_million: f64,
    pub use_streaming: bool,
    pub fallback_providers: Vec<FallbackProviderConfig>,
    pub credential_store_key: Option<String>,
    pub auth_method: String,
    pub api_key: Option<String>,
    pub retry: RetryConfig,
}

LLM provider configuration.

Fields

provider: String

Provider name: “openai”, “anthropic”, “local”.

model: String

Model identifier (e.g., “gpt-4o”, “claude-sonnet-4-20250514”).

api_key_env: String

Environment variable name containing the API key.

base_url: Option<String>

Optional base URL override for the API endpoint.

max_tokens: usize

Maximum tokens to generate in a response.

temperature: f32

Default temperature for generation.

context_window: usize

Context window size for the model.

input_cost_per_million: f64

Cost per 1M input tokens (USD).

output_cost_per_million: f64

Cost per 1M output tokens (USD).

use_streaming: bool

Whether to use streaming for LLM responses (enables token-by-token output).

fallback_providers: Vec<FallbackProviderConfig>

Optional fallback providers tried in order if the primary fails.

credential_store_key: Option<String>

Optional credential store key (provider name in the OS credential store). If set, the API key is loaded from the credential store instead of the env var.

auth_method: String

Authentication method: “api_key” (default) or “oauth”. When set to “oauth”, the provider will use an OAuth token from the credential store instead of a traditional API key.

api_key: Option<String>

Optional direct API key value. If the value starts with “keychain:”, the remainder is used as a keyring service name and the actual key is resolved at startup via resolve_credentials().

retry: RetryConfig

Retry configuration for transient API errors (429, 5xx, timeouts).

Implementations

impl LlmConfig

pub fn validate(&self) -> Vec<String>

Validate this LLM config and return any warnings.

Returns an empty Vec if the config is valid. Returns human-readable warning messages for problematic values (backward compatible — does not error).

Trait Implementations

impl Clone for LlmConfig

fn clone(&self) -> LlmConfig

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for LlmConfig

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for LlmConfig

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for LlmConfig

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for LlmConfig

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.