List your company

Merge Conflict Resolution on Merge-Bench Java (test)

54.7Equivalent Text RateGemini 2.5 Pro
-2.18812.58127.3542.119May 25, 2026

Evaluation results

MethodLinks
Gemini 2.5 Pro2026.0554.762.534.13.40
LLMergeJParameters=14B2026.0548.858.935.55.60
o3 Pro2026.0546.154.335.110.60
Claude Opus 42026.0544.451.227.621.20
Qwen3Parameters=14B, Training=SFT2026.0536.644.736.119.20
R1-0528Parameters=671B2026.0535.942.42433.20.4
Grok 42026.0533.439.717.942.40
QwQParameters=32B2026.0532.143.230.920.55.5
Qwen3Parameters=235B2026.0530.939.525.435.10
Llama 4 Maverick2026.0526.232.627.140.10.2
R1Parameters=70B2026.0525.73326.939.60.5
R1Parameters=32B2026.0522.830.429.339.70.6
Qwen3Parameters=32B2026.0513.216.911.171.80.1
Qwen3Parameters=14B2026.0512.916.68.774.70
R1Parameters=14B2026.059.313.415.470.70.5
Qwen3Parameters=8B2026.055.59.14.786.10.1
R1Parameters=8B2026.053.58.131.658.41.9
R1Parameters=1.5B2026.0500.246.5449.3