Atlas

Benchmarks

← All benchmarks

LiveCodeBench (Vals Public) - Easy

Code · 2024-03-12

The easy difficulty slice of Vals AI's LiveCodeBench public-dataset evaluation surface.

Top models (higher is better)

ModelScore
Opus 499.1
Opus 4.599.1
Opus 4.799.1
Opus 4.899.1
Claude Opus 599.1
Sonnet 4.599.1
Gemini 3.1 Pro Preview99.1
Gemini 3.5 Flash99.1
Gemini 3.6 Flash99.1
Gemini 3 Flash Preview99.1
Gemini 3 Pro Preview99.1
GPT-5.2-Codex99.1
GPT-5.3-Codex99.1
GPT-5.499.1
GPT-5 Mini99.1
GPT-5 Nano99.1
Grok 499.1
Grok 4.2099.1
Grok 4.599.1
Kimi K2.599.1
Kimi K2.699.1
MiMo-V2.5-Pro99.1
Qwen3.6 Plus (2026-04-02)99.1
Qwen3 Max (rolling alias)99.1
Claude Fable 598.8
Opus 4.198.8
Sonnet 498.8
Sonnet 4.698.8
DeepSeek-V4-Pro98.8
Gemini 2.5 Pro Preview 03-2598.8
Gemini 3.1 Flash-Lite Preview98.8
Gemini 3.5 Flash-Lite98.8
GLM-598.8
GPT-5.1-Codex98.8
GPT-5.198.8
GPT-5.298.8
GPT-598.8
Grok 3 Mini98.8
Grok 4.1 Fast98.8
Kimi K398.8
Loading Atlas data…