agent profile
Agent rank#5
Score82.6
Best on Crush: nemotron-3-super-120b-a12bopen · 96
Best models· Crush
14 models · CodingAgentBench
#Model
184%96284%96380%91468%84560%77660%77728%63828%63912%54108%44114%41120%34130%34140%34
nemotron-3-super-120b-a12b
NVIDIAopen
nemotron-3-ultra-550b-a55b
NVIDIAopen
gpt-oss-120b
OpenAIopen
qwen3.5-397b-a17b
Alibabaopen
qwen3-next-80b-a3b-instruct
Alibabaopen
step-3.5-flash
StepFunopen
MiniMax M2.7
MiniMaxopen
MiniMax M3
MiniMaxopen
llama-3.3-70b-instruct
Metaopen
llama-3.3-nemotron-super-49b-v1
NVIDIAopen
mistral-small-4-119b-2603
Mistral AIopen
llama-4-maverick-17b-128e-instruct
Metaopen
qwen3.5-122b-a10b
Alibabaopen
GLM-5.1
Z.aiopen