Share your thoughts, 1 month free Claude Pro on usSee more

File-level localization on SWE-Bench Verified

89.6Hit@1

MiniMax-M2.5

Updated 1mo ago

Evaluation Results

Method	Links
MiniMax-M2.5 2026.06		89.6	86.4	88.8	86.3	489
MiniMax-M2.5 2026.06		88.6	84.8	88	85.1	461
Qwen3-235B-A22B-Thinking 2026.06		88.4	81.3	87.6	83.5	498
Qwen3-Coder-480B-A35B 2026.06		86.8	83.4	84.4	82.1	491
Qwen3-Coder-480B-A35B 2026.06		86.2	83.3	83.9	81.8	488
MiniMax-M2.5 2026.06		83.4	85.5	82.3	81.6	487
MiniMax-M2.5 2026.06		82.4	84.6	81.8	80.9	493
Qwen3-Coder-480B-A35B 2026.06		81.6	83.7	77.3	77.6	490
Qwen3-Coder-480B-A35B 2026.06		81.2	83.7	78.5	78.5	490
Qwen3-Coder-30B-A3B 2026.06		78.8	73.5	78.1	74.3	457
Qwen3-Coder-30B-A3B 2026.06		77.2	73	76.3	73.4	437
Qwen3-Next-80B-A3B-Instruct 2026.06		76.6	72.6	76.6	73.1	447
Qwen3-Next-80B-A3B-Instruct 2026.06		76	72.1	76.1	72.8	441
Qwen3-Coder-30B-A3B 2026.06		75.2	75.9	74.8	73	461
Qwen3-Next-80B-A3B-Instruct 2026.06		75.2	77.6	75.2	73.8	480
Qwen3-Next-80B-A3B-Instruct 2026.06		74.4	77.5	73.4	73.1	479
Qwen3-Coder-30B-A3B 2026.06		73.4	73.8	72.5	70.8	456
Qwen3-Coder-Next 2026.06		56.2	55.3	55.1	54.2	305
Qwen3-Coder-Next 2026.06		56.2	58.4	54.3	54.7	357
Qwen3-Coder-Next 2026.06		56.2	57.3	54	53.8	350
Qwen3-Coder-Next 2026.06		49.4	48.3	48.5	47.4	271