List your company
ResearchFormerly WizWand

Coding on LiveCodeBench v6 (Score (%))

92Score (%)Fugu-Ultra
6.09628.39850.773.002Apr 17, 2026Apr 27, 2026May 8, 2026May 19, 2026May 29, 2026Jun 9, 2026Jun 20, 2026
Updated 4mo ago

Evaluation results

MethodLinks
Fugu-Ultra2026.0692
GPT-5.52026.0690.7
Fugu2026.0690.3
Claude Opus 4.82026.0690.3
Gemini 3.12026.0688.9
Confident DecodingModel=gpt-oss-120B, Decoding Strategy=Confident (Ours)2026.0684.5
Last LayerModel=gpt-oss-120B, Decoding Strategy=Last Layer2026.0683.1
Confident DecodingModel=Qwen3.5-122B-A10B, Decoding Strategy=Confident (Ours)2026.0679.2
Confident DecodingModel=Gemma-4-31B, Decoding Strategy=Confident (Ours)2026.0678.3
Confident DecodingModel=gpt-oss-20b, Decoding Strategy=Confident (Ours)2026.0677.6
Last LayerModel=gpt-oss-20b, Decoding Strategy=Last Layer2026.0677.1
Last LayerModel=Qwen3.5-122B-A10B, Decoding Strategy=Last Layer2026.0676.8
Mellum 2 (SFT)Post-training Stage=SFT, Parameters=2.5B/12B, Variant=Thinking2026.0575.1
Last LayerModel=Gemma-4-31B, Decoding Strategy=Last Layer2026.0675.1
Confident DecodingModel=Qwen3.5-35B-A3B, Decoding Strategy=Confident (Ours)2026.0674.4
Confident DecodingModel=Qwen3.5-27B, Decoding Strategy=Confident (Ours)2026.0673.3
Last LayerModel=Qwen3.5-35B-A3B, Decoding Strategy=Last Layer2026.0670.1
Mellum 2 (RL)Post-training Stage=RL, Parameters=2.5B/12B, Variant=Thinking2026.0569.9
Qwen3.5-9BParameters=9B, Variant=Thinking2026.0568.3
ZAYA1-8BActive / Total Parameters=0.7B / 8.0B, Sampling Settings=T=0.6, top-p=0.95, top-k=202026.0564.8
Keye-VL-2.0-30B-A3BModel Size=30B-A3B2026.0664.2
Last LayerModel=Qwen3.5-27B, Decoding Strategy=Last Layer2026.0663.9
Qwen3.5Model Size=35B-A3B2026.0662.8
OLMo-3-7BParameters=7B, Variant=Thinking2026.0559.8
Qwen3.5-4BParameters=4B, Variant=Thinking2026.0559.4
Qwen3.5-4BActive / Total Parameters=4.0B / 4.0B, Sampling Settings=Recommended sampling settings2026.0555.8
Qwen3-4B-Thinking-2507Active / Total Parameters=4.0B / 4.0B, Sampling Settings=Recommended sampling settings2026.0554.9
Gemma-4-E4B-itActive / Total Parameters=4.0B / 8.0B, Sampling Settings=Recommended sampling settings2026.0554.2
GPT-5-miniRelease Date/Version=2025-08-072026.0651.5
Ministral-3-14BParameters=14B, Variant=Thinking2026.0542.7
Pair+Backbone=olmo-3.1-32b-it-sft2026.0429.7
DPOBackbone=olmo-3.1-32b-it-sft2026.0429
Mag+Backbone=olmo-3.1-32b-it-sft2026.0428.6
Soft+Backbone=olmo-3.1-32b-it-sft2026.0427.5
MPO+Backbone=olmo-3.1-32b-it-sft2026.0426.6
RFTBackbone=olmo-3.1-32b-it-sft2026.0425.3
Mag+Backbone=olmo-3-7b-it-sft2026.0422.5
BaseBackbone=olmo-3.1-32b-it-sft2026.0421.4
DPOBackbone=olmo-3-7b-it-sft2026.0419.5
MPO+Backbone=olmo-3-7b-it-sft2026.0419.5
RFTBackbone=olmo-3-7b-it-sft2026.0418.7
Pair+Backbone=olmo-3-7b-it-sft2026.0418
Soft+Backbone=olmo-3-7b-it-sft2026.0416.1
BaseBackbone=olmo-3-7b-it-sft2026.0414.9
Mag+Backbone=gemma-3-4b-sft2026.0411.5
RFTBackbone=gemma-3-4b-sft2026.0410.7
MPO+Backbone=gemma-3-4b-sft2026.0410.7
DPOBackbone=gemma-3-4b-sft2026.0410.6
Pair+Backbone=gemma-3-4b-sft2026.0410.2
Soft+Backbone=gemma-3-4b-sft2026.049.8
BaseBackbone=gemma-3-4b-sft2026.049.4