LiveBench Coding Completion — leaderboard

Metric: Score. Source: livebench.ai. 65 models tracked.

Top models

#ModelScore
1Claude 3.5 Sonnet (20240620)68.42
2GPT-4o (2024-08-06)55.26
3GPT-4o ChatGPT50
4DeepSeek V2.550
5Gemini 1.5 Pro (Preview 0827)50
6Llama 3.1 405B Instruct47.37
7GPT-4o Mini (2024-07-18)44.74
8GPT-4o (2024-05-13)44.74
9Mistral Large 2 (Jul)44.74
10Gemini 1.5 Pro (Preview 0801)44.74
11GPT-4 Turbo42.1
12Claude 3 Opus (20240229)42.1
13GPT-4 (0613)42.1
14GPT-4 Preview (0125)42.1
15DeepSeek Coder V242.1

Interactive version: aibenchmarks.dev/benchmark?slug=livebench-coding-completion · How the rankings work · Data refreshed daily, snapshot 2026-07-09.