BenchAtlas

Rankings / Z.ai

GLM 4.6

released 2025-09-30

BenchAtlas Index

as of 2026-07-14
46.5
thinking
rank #246
10 families · 5 categories · high
45.5
base configuration
rank #253
6 families · 3 categories · medium
35.0
no reasoning
rank #324
10 families · 5 categories · high

Benchmark evidence

60 results
agentic coding
τ²-Benchno reasoning76.9 %independentArtificial Analysis
τ²-Benchthinking70.5 %independentArtificial Analysis
τ²-Bench
subset=banking
thinking10.5 %independentArtificial Analysis
coding
LiveCodeBench v6
implementation=artificial-analysis
thinking69.5 %independentArtificial Analysis
LiveCodeBench v6
implementation=artificial-analysis
no reasoning56.1 %independentArtificial Analysis
SciCodethinking38.4 %independentArtificial Analysis
SciCodeno reasoning33.1 %independentArtificial Analysis
external indices
AA Coding Indexthinking45.8 pointsindependentArtificial Analysis
Intelligence Index v4.1thinking28.7 pointsindependentArtificial Analysis
Intelligence Index v4.1no reasoning23.0 pointsindependentArtificial Analysis
AA Math Indexthinking86.0 pointsindependentArtificial Analysis
AA Math Indexno reasoning44.3 pointsindependentArtificial Analysis
ECI140.9 pointsindependentEpoch AI Benchmarking Hub
human preference
Chinese (style control)older version
arena=text · category=chinese · style_control=true
1482.5
1467.8–1497.2
communityLMArena Leaderboard Dataset
Coding (style control)older version
arena=text · category=coding · style_control=true
1459.0
1451.8–1466.3
communityLMArena Leaderboard Dataset
Industry Software And It Services (style control)older version
arena=text · category=industry_software_and_it_services · style_control=true
1453.7
1447.8–1459.6
communityLMArena Leaderboard Dataset
Industry Medicine And Healthcare (style control)older version
arena=text · category=industry_medicine_and_healthcare · style_control=true
1450.4
1436.5–1464.3
communityLMArena Leaderboard Dataset
Hard Prompts English (style control)older version
arena=text · category=hard_prompts_english · style_control=true
1448.0
1441.5–1454.6
communityLMArena Leaderboard Dataset
French (style control)older version
arena=text · category=french · style_control=true
1447.3
1422.7–1471.8
communityLMArena Leaderboard Dataset
Industry Life And Physical And Social Science (style control)older version
arena=text · category=industry_life_and_physical_and_social_science · style_control=true
1447.2
1439.1–1455.2
communityLMArena Leaderboard Dataset
Hard Prompts (style control)older version
arena=text · category=hard_prompts · style_control=true
1442.1
1437.1–1447.2
communityLMArena Leaderboard Dataset
Expert (style control)older version
arena=text · category=expert · style_control=true
1441.4
1427.8–1455.0
communityLMArena Leaderboard Dataset
Industry Mathematical (style control)older version
arena=text · category=industry_mathematical · style_control=true
1440.7
1426.5–1454.9
communityLMArena Leaderboard Dataset
Industry Legal And Government (style control)older version
arena=text · category=industry_legal_and_government · style_control=true
1436.8
1424.8–1448.9
communityLMArena Leaderboard Dataset
English (style control)older version
arena=text · category=english · style_control=true
1436.7
1431.5–1441.9
communityLMArena Leaderboard Dataset
Longer Query (style control)older version
arena=text · category=longer_query · style_control=true
1432.8
1426.1–1439.5
communityLMArena Leaderboard Dataset
German (style control)older version
arena=text · category=german · style_control=true
1430.2
1407.1–1453.4
communityLMArena Leaderboard Dataset
Overall (style control)
arena=text · category=overall · style_control=true
1425.3
1421.4–1429.2
communityLMArena Leaderboard Dataset
Multi Turn (style control)older version
arena=text · category=multi_turn · style_control=true
1421.3
1413.2–1429.5
communityLMArena Leaderboard Dataset
Math (style control)older version
arena=text · category=math · style_control=true
1420.6
1407.8–1433.3
communityLMArena Leaderboard Dataset
Exclude Ties (style control)older version
arena=text · category=exclude_ties · style_control=true
1417.0
1411.4–1422.5
communityLMArena Leaderboard Dataset
Polish (style control)older version
arena=text · category=polish · style_control=true
1416.6
1399.6–1433.5
communityLMArena Leaderboard Dataset
Instruction Following (style control)older version
arena=text · category=instruction_following · style_control=true
1415.3
1408.9–1421.7
communityLMArena Leaderboard Dataset
Spanish (style control)older version
arena=text · category=spanish · style_control=true
1414.6
1394.0–1435.3
communityLMArena Leaderboard Dataset
Industry Business And Management And Financial Operations (style control)older version
arena=text · category=industry_business_and_management_and_financial_operations · style_control=true
1412.9
1405.3–1420.5
communityLMArena Leaderboard Dataset
Russian (style control)older version
arena=text · category=russian · style_control=true
1410.2
1399.2–1421.2
communityLMArena Leaderboard Dataset
Non English (style control)older version
arena=text · category=non_english · style_control=true
1409.4
1404.3–1414.4
communityLMArena Leaderboard Dataset
Industry Writing And Literature And Language (style control)older version
arena=text · category=industry_writing_and_literature_and_language · style_control=true
1405.1
1398.1–1412.1
communityLMArena Leaderboard Dataset
Creative Writing (style control)older version
arena=text · category=creative_writing · style_control=true
1402.3
1393.7–1410.9
communityLMArena Leaderboard Dataset
Industry Entertainment And Sports And Media (style control)older version
arena=text · category=industry_entertainment_and_sports_and_media · style_control=true
1394.8
1387.2–1402.4
communityLMArena Leaderboard Dataset
Korean (style control)older version
arena=text · category=korean · style_control=true
1365.2
1339.5–1390.9
communityLMArena Leaderboard Dataset
knowledge science
GPQA Diamond
implementation=artificial-analysis
thinking78.0 %independentArtificial Analysis
GPQA Diamond
implementation=vals-ai
74.5 %independentVals AI
GPQA Diamond
implementation=artificial-analysis
no reasoning63.2 %independentArtificial Analysis
Humanity's Last Exam
implementation=artificial-analysis
thinking13.3 %independentArtificial Analysis
Humanity's Last Exam
implementation=artificial-analysis
no reasoning5.2 %independentArtificial Analysis
MMLU-Pro
implementation=artificial-analysis
thinking82.9 %independentArtificial Analysis
MMLU-Pro
implementation=vals-ai
82.2 %independentVals AI
MMLU-Pro
implementation=artificial-analysis
no reasoning78.4 %independentArtificial Analysis
long context instruction
AA-LCRthinking54.3 %independentArtificial Analysis
AA-LCRno reasoning26.3 %independentArtificial Analysis
IFBenchthinking43.4 %independentArtificial Analysis
IFBenchno reasoning36.7 %independentArtificial Analysis
professional
CorpFin56.8 %independentVals AI
LegalBench79.6 %independentVals AI
MedQA92.2 %independentVals AI
TaxEval66.2 %independentVals AI
reasoning math
AIME
implementation=vals-ai
92.7 %independentVals AI
AIME
year=2025 · implementation=artificial-analysis
thinking86.0 %independentArtificial Analysis
AIME
year=2025 · implementation=artificial-analysis
no reasoning44.3 %independentArtificial Analysis

Agent + model results

systems, not bare-model scores
agent + model GLM-4.6 + GLM 4.6SWE-bench Verified68.2 %unverifiedSWE-bench Leaderboard
agent + model mini-SWE-agent + GLM 4.6SWE-bench Verified55.4 %communitySWE-bench Leaderboard
agent + model mini-SWE-agent + GLM 4.6SWE-bench bash-only55.4 %communitySWE-bench Leaderboard
agent + model mini-SWE-agent + GLM 4.6SWE-bench Verified0.1 usd_per_taskcommunitySWE-bench Leaderboard
agent + model mini-SWE-agent + GLM 4.6SWE-bench bash-only0.1 usd_per_taskcommunitySWE-bench Leaderboard
agent + model Artificial Analysis harness + GLM 4.6Terminal-Bench 2.149.4 %independentArtificial Analysis
agent + model Artificial Analysis harness + GLM 4.6Terminal-Bench Hard28.8 %independentArtificial Analysis
agent + model Artificial Analysis harness + GLM 4.6Terminal-Bench Hard25.0 %independentArtificial Analysis

These scores measure the whole agent system (scaffold, tools, budgets) — they are never merged into the bare model’s numbers.

Related variants

GLM 5GLM 5.2GLM 5.1GLM 4.7GLM 4.5GLM 4 6vGLM 4 5vGLM 4 0520