Struct LanguageModel 

pub struct LanguageModel {
    pub name: String,
    pub aliases: Vec<String>,
    pub version: String,
    pub input_modalities: Vec<Modality>,
    pub output_modalities: Vec<Modality>,
    pub prompt_text_token_price: i64,
    pub prompt_image_token_price: i64,
    pub cached_prompt_token_price: i64,
    pub completion_text_token_price: i64,
    pub search_price: i64,
    pub max_prompt_length: i32,
    pub system_fingerprint: String,
}

Information about a language model.

This struct contains comprehensive metadata about an xAI language model, including its capabilities, pricing, and technical specifications.

Pricing Units

The pricing fields use specific units to represent fractional cents:

• prompt_text_token_price: 1/100 USD cents per 1M tokens (e.g., 500 = $0.05 per 1M tokens)
• prompt_image_token_price: 1/100 USD cents per 1M tokens
• completion_text_token_price: 1/100 USD cents per 1M tokens
• cached_prompt_token_price: USD cents per 100M tokens (e.g., 50 = $0.50 per 100M tokens)
• search_price: 1/100 USD cents per 1M searches

Use calculate_cost to convert these to USD amounts.

Examples

let model = client.get_model("grok-2-1212").await?;

// Check capabilities
if model.supports_multimodal() {
    println!("{} supports images!", model.name);
}

// Calculate costs
let cost = model.calculate_cost(50_000, 5_000, 0);
println!("50K prompt + 5K completion costs: ${:.4}", cost);

Fields

name: String

The model name used in API requests (e.g., “grok-2-1212”).

aliases: Vec<String>

Alternative names that can be used for this model (e.g., [“grok-2-latest”]).

Aliases provide convenient shortcuts for referring to models without needing to know the specific version number.

version: String

Version number of the model (e.g., “2.0”).

input_modalities: Vec<Modality>

Supported input modalities.

Common combinations:

• [Text] - Text-only model
• [Text, Image] - Multimodal model supporting vision

output_modalities: Vec<Modality>

Supported output modalities.

Most models output [Text], but some specialized models may support image generation or embeddings.

prompt_text_token_price: i64

Price per million prompt text tokens in 1/100 USD cents.

Example: 500 = $0.05 per 1M tokens = $0.00005 per token

prompt_image_token_price: i64

Price per million prompt image tokens in 1/100 USD cents.

Only applicable for multimodal models that accept images.

cached_prompt_token_price: i64

Price per 100 million cached prompt tokens in USD cents.

Example: 50 = $0.50 per 100M tokens

Cached tokens are significantly cheaper as they’re reused from previous requests with the same prefix.

completion_text_token_price: i64

Price per million completion text tokens in 1/100 USD cents.

Example: 1500 = $0.15 per 1M tokens = $0.00015 per token

search_price: i64

Price per million searches in 1/100 USD cents.

Only applicable when using web search or X search tools.

max_prompt_length: i32

Maximum context length in tokens (prompt + completion).

This represents the total number of tokens the model can process in a single request, including both input and output.

system_fingerprint: String

Backend configuration fingerprint.

This identifier tracks the specific backend configuration used by the model, useful for debugging and reproducibility.

Implementations

impl LanguageModel

pub fn calculate_cost( &self, prompt_tokens: u32, completion_tokens: u32, cached_tokens: u32, ) -> f64

Calculate the cost (in USD) for a given number of prompt and completion tokens.

Examples

let cost = model.calculate_cost(1000, 500, 0);
println!("Cost: ${:.4}", cost);

pub fn supports_multimodal(&self) -> bool

Check if the model supports multimodal input (text + images).

Returns true if the model accepts both text and image inputs, allowing you to send image URLs alongside text prompts.

Examples

let model = client.get_model("grok-2-vision-1212").await?;

if model.supports_multimodal() {
    println!("{} can process images!", model.name);
} else {
    println!("{} is text-only", model.name);
}

Trait Implementations

impl Clone for LanguageModel

fn clone(&self) -> LanguageModel

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for LanguageModel

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

Formats the value using the given formatter.