Crate llm_models_spider 

llm_models_spider

Auto-updated registry of LLM model capabilities, rankings, and pricing.

This crate provides zero-cost compile-time lookups for model capabilities like vision support, audio input, etc. The model lists are automatically updated via GitHub Actions by scraping OpenRouter, LiteLLM, and Chatbot Arena.

Usage

use llm_models_spider::{supports_vision, supports_audio, ModelCapabilities};

// Quick checks
assert!(supports_vision("gpt-4o"));
assert!(supports_vision("claude-3-sonnet"));
assert!(!supports_vision("gpt-3.5-turbo"));

// Get full capabilities
if let Some(caps) = ModelCapabilities::lookup("google/gemini-2.0-flash") {
    println!("Vision: {}, Audio: {}", caps.vision, caps.audio);
}

Rich Model Profiles

use llm_models_spider::{model_profile, arena_rank};

if let Some(profile) = model_profile("gpt-4o") {
    println!("Max input: {} tokens", profile.max_input_tokens);
    if let Some(cost) = profile.pricing.input_cost_per_m_tokens {
        println!("Input cost: ${}/M tokens", cost);
    }
}

if let Some(rank) = arena_rank("gpt-4o") {
    println!("Arena rank: {:.1}/100", rank);
}

Auto-Updates

Model data is fetched from OpenRouter, LiteLLM, and Chatbot Arena, then committed to this repo via scheduled GitHub Actions. New releases are published automatically when the data changes.

Structs

ModelCapabilities

Model capabilities struct.

ModelInfoEntry

Detailed model information entry.

ModelPricing

Pricing in USD.

ModelProfile

Full model profile: capabilities, ranks, pricing, and context window.

ModelRanks

Arena and task-specific rankings, normalized to 0.0-100.0.

Constants

ALL_MODELS

All known models (union of all lists).

AUDIO_MODELS

Models that support audio input.

MODEL_INFO

Sorted by name for binary search lookup.

TEXT_ONLY_MODELS

Models that are text-only (no vision/audio support).

VISION_MODELS

Models that support vision/image input.

Functions

arena_rank

Get the arena rank (0.0-100.0) for a model, if available.

is_text_only

Check if a model is text-only (no vision/audio).

model_profile

Get a full model profile including capabilities, pricing, ranks, and context window.

supports_audio

Check if a model supports audio input.

supports_pdf

Check if a model supports PDF/file input.

supports_video

Check if a model supports video input.

supports_vision

Check if a model supports vision/image input.