DataLearner logo

GLM-4.6 Benchmark Details

GLM-4.6 currently shows benchmark results led by AIME2025 (16 / 110, score 98.60), LiveCodeBench (22 / 125, score 84.50), MMLU-Pro (47 / 175, score 83).

Benchmark Results

GLM-4.6

Benchmark Results

Thinking
Tool usage

Knowledge Exams

5 evaluations
Benchmark / mode
Score
Rank/total
MMLU-Pro
Standard Mode
78
74 / 175
MMLU-Pro
Thinking Mode
83
47 / 175
HLE
Standard Mode
5.20
224 / 235
HLE
Thinking Mode
17.20
174 / 235
HLE
Thinking ModeTools
30.40
117 / 235

Scientific Reasoning

4 evaluations
Benchmark / mode
Score
Rank/total
GPQA Diamond
Standard Mode
63
206 / 253
GPQA Diamond
Thinking Mode
81
122 / 253
GPQA Diamond
Thinking ModeTools
82.90
108 / 253
CritPt
Thinking Mode
1.10
151 / 204

Repository Engineering

2 evaluations
Benchmark / mode
Score
Rank/total
SWE-bench Verified
Standard Mode
68
71 / 116
SWE-bench Verified
Thinking ModeTools
68
71 / 116

Algorithmic Coding

2 evaluations
Benchmark / mode
Score
Rank/total
LiveCodeBench
Thinking Mode
82.80
28 / 125
LiveCodeBench
Thinking ModeTools
84.50
22 / 125

Mathematics

4 evaluations
Benchmark / mode
Score
Rank/total
AIME2025
Standard Mode
44
101 / 110
AIME2025
Thinking Mode
98.60
16 / 110
AIME2025
Thinking ModeTools
98.60
16 / 110
FrontierMath - Tier 4
Standard Mode
2.10
56 / 80

Writing

1 evaluations
Benchmark / mode
Score
Rank/total
Creative Writing
Standard Mode
1411.10
68 / 110

Agentic Development

3 evaluations
Benchmark / mode
Score
Rank/total
Terminal-Bench
Thinking ModeTools
40.50
12 / 35
Terminal Bench Hard
Standard ModeTools
28.80
116 / 244
Terminal Bench Hard
Thinking ModeTools
25
128 / 244

Service Workflows

3 evaluations
Benchmark / mode
Score
Rank/total
τ²-Bench
Thinking ModeTools
75.90
21 / 44
τ²-Bench - Telecom
Thinking ModeTools
71
23 / 30
τ³-Banking
Thinking ModeTools
13.40
119 / 167

Instruction Following

1 evaluations
Benchmark / mode
Score
Rank/total
IF Bench
Thinking Mode
43
31 / 31

Fact Finding

1 evaluations
Benchmark / mode
Score
Rank/total
BrowseComp
Thinking ModeTools
45.10
48 / 58

Cross-capability Suites

1 evaluations
Benchmark / mode
Score
Rank/total
LiveBench
Standard Mode
55.19
83 / 117
Compare with other models