AI Model Release Tracker
Competitive coding

LiveCodeBench

Coding problems published so recently the AI can't have seen them in training — a contamination-free test of raw programming skill. Higher is better.

Rankings

Higher is better
#1
DeepSeek-V4-Pro
93.5%
2
DeepSeek-V4-Flash
91.6%
2
Qwen3.7-Max
91.6%
4
Qwen3.8-27BOW
90.3%
5
Kimi K2.6
89.6%
5
Qwen3.7-Plus
89.6%
7
Kimi K2.5
85%
8
GLM-4.7
84.9%

LiveCodeBench — frequently asked questions

What is LiveCodeBench?
Coding problems published so recently the AI can't have seen them in training — a contamination-free test of raw programming skill. Higher is better.
Which AI model scores highest on LiveCodeBench?
DeepSeek-V4-Pro by DeepSeek holds the best LiveCodeBench result among tracked models, at 93.5% (released Apr 24 2026). Higher scores are better on this benchmark.
What are the top 5 models on LiveCodeBench?
1. DeepSeek-V4-Pro (DeepSeek) — 93.5%; 2. DeepSeek-V4-Flash (DeepSeek) — 91.6%; 2. Qwen3.7-Max (Qwen) — 91.6%; 4. Qwen3.8-27B (Qwen) — 90.3%; 5. Kimi K2.6 (Moonshot AI) — 89.6%.
How many models have a published LiveCodeBench score?
8 tracked models have a published LiveCodeBench score. Scores are the figures reported by each lab at that model's release, so this page is a record of results over time rather than a re-run leaderboard.
What is the best open model on LiveCodeBench?
Qwen3.8-27B by Qwen is the highest-ranked model with downloadable weights on LiveCodeBench, scoring 90.3% at rank 4 overall.