Coding benchmark

LiveCodeBench leaderboard

7 models have a published score. DeepSeek V4 Pro (DeepSeek) leads with 93.5, and the spread to the lowest score, Gemma 4 26B A4B at 77.1, is 16.4 points. The highest-scoring model you can deploy on OpenLLM Buddy is Qwen3.8 27B, ranked 2 of 7 with 90.3. Sources report different settings for this test, such as test year, version, or reasoning effort. Each setting is listed in the source table, so compare close scores with care.

Best openllm buddy catalog

90.3

Qwen3.8 27B

4 models scored

Best open weights

93.5

DeepSeek V4 Pro

3 models scored

Best closed api

No published score

0 models scored

LiveCodeBench scores

Pass@1: the generated program must pass every hidden test case. The problem window is noted where a source gives it.

  • OpenLLM Buddy catalog
  • Open weights
  • Closed API
  1. 1DeepSeek V4 ProV4-Pro (Preview) Max, version not stated93.5
  2. 2Qwen3.8 27BLiveCodeBench v690.3
  3. 3Nemotron 3 UltraLiveCodeBench v689.0
  4. 4Reflection Beam 1.0BF16, version not stated88.3
  5. 5Qwen3.6 27BLiveCodeBench v683.9
  6. 6Gemma 4 31BLiveCodeBench v680.0
  7. 7Gemma 4 26B A4BLiveCodeBench v677.1

What it measures

Problems are pulled continuously from LeetCode, AtCoder, and Codeforces and tagged with their release date, so a model can be scored only on problems it could not have seen in training.

How it is scored

Pass@1: the generated program must pass every hidden test case. The problem window is noted where a source gives it.

Why it matters

Its rolling, dated problem set makes it one of the most contamination-resistant coding benchmarks.

Official benchmark page

More coding tests

Sources

ModelTestScoreSettingSource
DeepSeek V4 ProLiveCodeBench93.5V4-Pro (Preview) Max, version not statedHugging Face model card (DeepSeek-V4-Pro)
Qwen3.8 27BLiveCodeBench90.3LiveCodeBench v6Qwen3.8-27B model card (Hugging Face)
Nemotron 3 UltraLiveCodeBench89.0LiveCodeBench v6Hugging Face model card (NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16)
Reflection Beam 1.0LiveCodeBench88.3BF16, version not statedHugging Face model card (Beam-1.0-501B-A23B)
Qwen3.6 27BLiveCodeBench83.9LiveCodeBench v6Qwen3.6-27B model card (Hugging Face)
Gemma 4 31BLiveCodeBench80.0LiveCodeBench v6Gemma 4 31B IT model card (Hugging Face)
Gemma 4 26B A4BLiveCodeBench77.1LiveCodeBench v6Gemma 4 31B IT model card (Hugging Face), 26B A4B column

Scores are published results from model cards, lab reports, and public leaderboards, each linked to its source. Settings such as thinking mode or test version differ between sources and are shown next to each score. OpenLLM Buddy catalog models are uncensored fine-tunes; their scores are those of the official base model. Data collected 2026-10-30. How the index works