Struct LlmConfig 

pub struct LlmConfig {
    pub model: String,
    pub api_key: Option<String>,
    pub base_url: Option<String>,
    pub temp: f64,
    pub max_tokens: Option<u32>,
    pub prompt_cache_key: Option<String>,
    pub project_id: Option<String>,
    pub location: Option<String>,
}

LLM provider configuration — single config for any provider.

Two optional fields control routing:

• api_key: None → auto from env vars (OPENAI_API_KEY, ANTHROPIC_API_KEY, etc.)
• base_url: None → auto-detect provider from model name; Some → custom endpoint

use sgr_agent::LlmConfig;

let c = LlmConfig::auto("gpt-4o");                                          // env vars
let c = LlmConfig::with_key("sk-...", "claude-3-haiku");                    // explicit key
let c = LlmConfig::endpoint("sk-or-...", "https://openrouter.ai/api/v1", "gpt-4o"); // custom
let c = LlmConfig::auto("gpt-4o").temperature(0.9).max_tokens(2048);        // builder

Fields

model: String

api_key: Option<String>

base_url: Option<String>

temp: f64

max_tokens: Option<u32>

prompt_cache_key: Option<String>

OpenAI prompt cache key — caches system prompt prefix server-side.

project_id: Option<String>

Vertex AI project ID (enables Vertex routing when set).

location: Option<String>

Vertex AI location (default: “global”).

Implementations

impl LlmConfig

pub fn auto(model: impl Into<String>) -> Self

Auto-detect provider from model name, use env vars for auth.

pub fn with_key(api_key: impl Into<String>, model: impl Into<String>) -> Self

Explicit API key, auto-detect provider from model name.

pub fn endpoint( api_key: impl Into<String>, base_url: impl Into<String>, model: impl Into<String>, ) -> Self

Custom OpenAI-compatible endpoint (OpenRouter, Ollama, LiteLLM, etc.).

pub fn vertex(project_id: impl Into<String>, model: impl Into<String>) -> Self

Vertex AI — uses gcloud ADC for auth (no API key needed).

pub fn location(self, loc: impl Into<String>) -> Self

Set Vertex AI location.

pub fn temperature(self, t: f64) -> Self

Set temperature.

pub fn max_tokens(self, m: u32) -> Self

Set max output tokens.

pub fn prompt_cache_key(self, key: impl Into<String>) -> Self

Set OpenAI prompt cache key for server-side system prompt caching.

pub fn label(&self) -> String

Human-readable label for display.

pub fn compaction_model(&self) -> String

Infer a cheap/fast model for compaction based on the primary model.

pub fn for_compaction(&self) -> Self

Create a compaction config — cheap model, low max_tokens.

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.