LMSys Coding Leaderboard

A mix indexed on the top coding models in the LMSYS leaderboard. Models with a higher score and more consistent votes will be selected more often.

Updated Aug 13$4.24/M tokens$14.44/M tokensGithubGitHub

API

Example

import OpenAI from 'openai'


const openai = new OpenAI({
  baseURL: "https://api.crosshatch.app/v1",
  apiKey: "your_api_key_here"
})


const completion = await openai.chat.completions({
  model: "lmsys-coding",
  messages: [
    {
      role: "user",
      content: "Hello there"
    }
  ]
})

Models

This mix uses the models below:

OpenAI Icon

chatgpt-4o-latest

Provided by multiple sources

Anthropic Icon

claude-3-5-sonnet-20240620

Provided by multiple sources

OpenAI Icon

gpt-4o-2024-05-13

Provided by multiple sources

Google Gemini Icon

gemini-1.5-pro-exp-0827

Provided by multiple sources

Readme

LMSys Coding Leaderboard Mix

Don't worry about staying up to date with the latest coding model to use in your favorite IDE. This mix is based on the top coding models ranked in the LMSys Chatbot Arena for the Coding category. The weight is a function of the Elo score adjusted for number of votes and variance. Models with more consistently high votes will be weighed more heavily.

Composition

This mix produces responses from the following models:

Model NameWeight %
chatgpt-4o-latest22.32%
claude-3-5-sonnet-2024062026.61%
gpt-4o-2024-05-1330.58%
gemini-1.5-pro-exp-082720.49%

Update Frequency: Based on LMSys (every few weeks)