Documentation Index
Fetch the complete documentation index at: https://mintlify.com/vibheksoni/t3router/llms.txt
Use this file to discover all available pages before exploring further.
Overview
The Config struct controls request parameters for the t3.chat API, including search functionality and reasoning effort levels.
Struct Definition
#[derive(Clone)]
pub struct Config {
pub include_search: bool,
pub reasoning_effort: ReasoningEffort,
}
Fields
Whether to include search functionality in the API request. When enabled, the assistant can search for information online.Default: false
The level of reasoning effort to use for the request. Higher levels provide more thorough responses but may take longer.Default: ReasoningEffort::Low
ReasoningEffort Enum
Controls the reasoning effort level for API requests.
#[derive(Clone, Copy)]
pub enum ReasoningEffort {
Low,
Medium,
High,
}
Variants
Minimal reasoning effort for faster responsesString value: "low"
Balanced reasoning effort for moderate complexityString value: "medium"
Maximum reasoning effort for complex queriesString value: "high"
as_str
Returns the string value for the reasoning effort.
pub fn as_str(&self) -> &'static str
The corresponding string value (“low”, “medium”, or “high”)
use t3router::t3::config::{Config, ReasoningEffort};
let effort = ReasoningEffort::High;
assert_eq!(effort.as_str(), "high");
Constructor
new
Creates a new Config instance with default values.
A new configuration object with default values:
include_search: falsereasoning_effort: ReasoningEffort::Low
use t3router::t3::config::Config;
let config = Config::new();
assert_eq!(config.include_search, false);
Usage Examples
Default configuration
use t3router::{Client, Message, Type, Config};
let mut client = Client::new(
"cookies".to_string(),
"session_id".to_string()
);
let msg = Message::new(Type::User, "Hello!".to_string());
// Use default config
let response = client.send("gpt-4", Some(msg), Some(Config::new())).await?;
Custom configuration with search enabled
use t3router::t3::{client::Client, message::{Message, Type}, config::{Config, ReasoningEffort}};
let mut config = Config::new();
config.include_search = true;
config.reasoning_effort = ReasoningEffort::Medium;
let msg = Message::new(
Type::User,
"What are the latest developments in quantum computing?".to_string()
);
let response = client.send("gpt-4", Some(msg), Some(config)).await?;
High reasoning effort for complex queries
use t3router::t3::config::{Config, ReasoningEffort};
let mut config = Config::new();
config.reasoning_effort = ReasoningEffort::High;
config.include_search = false;
let msg = Message::new(
Type::User,
"Explain the proof of Fermat's Last Theorem".to_string()
);
let response = client.send("gpt-4", Some(msg), Some(config)).await?;
Comparing reasoning levels
use t3router::t3::config::{Config, ReasoningEffort};
// Quick response
let low_config = Config {
include_search: false,
reasoning_effort: ReasoningEffort::Low,
};
// Balanced approach
let medium_config = Config {
include_search: true,
reasoning_effort: ReasoningEffort::Medium,
};
// Deep analysis
let high_config = Config {
include_search: true,
reasoning_effort: ReasoningEffort::High,
};
Best Practices
When to use different reasoning efforts
- Low: Simple queries, greetings, basic Q&A
- Medium: Moderate complexity, analysis tasks, code explanation
- High: Complex reasoning, mathematical proofs, detailed research
Search functionality
Enable include_search when:
- Asking about current events
- Requesting recent information
- Needing factual verification
- Researching specific topics
use t3router::t3::config::Config;
// For current events
let mut config = Config::new();
config.include_search = true;
let msg = Message::new(
Type::User,
"What happened in the news today?".to_string()
);
Omitting configuration
You can omit the config parameter entirely, and the default configuration will be used automatically:
// These are equivalent:
let response1 = client.send("gpt-4", Some(msg), None).await?;
let response2 = client.send("gpt-4", Some(msg), Some(Config::new())).await?;