benchgap
Zhipu (GLM) · model

GLM-5.3-Flash benchmark scores

As of 2026-11-21, GLM-5.3-Flash (Zhipu (GLM)) has measured scores on 43 benchmarks and estimated scores on 113 more.

BenchmarkScoreSource
AA-GPQA Diamond86.7%estimated ± 3.6 pp, high confidence
AA-HLE29.9%estimated ± 6.2 pp, medium confidence
AA-SciCode54.8%estimated ± 2.5 pp, high confidence
AIME2695.9%estimated ± 1.2 pp, medium confidence
Artificial Analysis Intelligence Index25.7%estimated ± 5.8 pp, medium confidence
CritPt5.9%estimated ± 5.3 pp, medium confidence
GDPval-AA63.0%estimated ± 5.6 pp, medium confidence
GPQA Diamond88.8%estimated ± 1.8 pp, high confidence
HLE40.1%estimated ± 9.2 pp, low confidence
LiveCodeBench v694.0%estimated ± 1.9 pp, medium confidence
AA-Omniscience Accuracy36.7%estimated ± 6.5 pp, medium confidence
SciCode37.8%estimated ± 2.9 pp, low confidence
Terminal-Bench 2.184.3%measured
BrowseComp89.0%estimated ± 4.0 pp, high confidence
HLE w/ tools55.3%measured
τ²-bench results97.9%estimated ± 3.1 pp, high confidence
AA Coding Index73.1%estimated ± 1.2 pp, high confidence
AA-MMMU-Pro79.1%estimated ± 2.3 pp, high confidence
APEX-Agents37.6%estimated ± 7.4 pp, low confidence
SWE-bench Verified81.6%estimated ± 2.3 pp, high confidence
AutomationBench48.8%measured
CyberGym84.1%estimated ± 3.1 pp, low confidence
SWE-bench Pro62.1%estimated ± 3.5 pp, high confidence
WideResearch85.9%estimated ± 3.9 pp, low confidence
AA Agentic Index56.2%estimated ± 3.3 pp, medium confidence
MMLU-Pro85.5%estimated ± 4.8 pp, high confidence
FailureSensorIQ64.9%estimated ± 2.1 pp, high confidence
GPQA88.1%estimated ± 4.5 pp, high confidence
MMLU93.4%estimated ± 1.2 pp, medium confidence
Vals GPQA Diamond86.4%measured
Vals LiveCodeBench80.5%measured
Vals MMLU-Pro86.1%measured
MMMU85.1%estimated ± 2.9 pp, low confidence
Chess Puzzles21.0%measured
AA MMLU-Pro87.2%estimated ± 1.1 pp, high confidence
Gert Labs77.5%estimated ± 9.1 pp, low confidence
AA-AnalystAgent46.1%estimated ± 5.8 pp, low confidence
AA EnterpriseOps-Gym47.0%estimated ± 7.2 pp, low confidence
AA Harvey LAB v1.089.6%estimated ± 1.7 pp, medium confidence
AA Tau3 Banking47.2%measured
AA Terminal-Bench 2.184.3%measured
ApprenticeBench13.1%estimated ± 12.5 pp, low confidence
ARC-AGI-180.7%estimated ± 6.5 pp, medium confidence
ARC-AGI-243.6%estimated ± 13.8 pp, low confidence
BabaIsBench67.0%measured
CursorBench 3.166.4%estimated ± 5.8 pp, low confidence
CursorBench 3.263.7%estimated ± 2.0 pp, high confidence
EBR-bench0.0%estimated ± 10.4 pp, low confidence
Terminal-Bench 3.029.5%estimated ± 3.0 pp, high confidence
FrontierCode 1.1 Main51.1%estimated ± 3.0 pp, medium confidence
FrontierMath v2 (Tier 4)17.1%measured
FrontierMath v2 (Tiers 1-3)55.8%measured
FrontierSWE v218.1%measured
Furniture Assembly32.9%estimated ± 9.3 pp, medium confidence
MaxxBench55.1%measured
MazeBench0.0%measured
Mystery Game Puzzles8.0%measured
NYT Connections77.8%estimated ± 6.4 pp, low confidence
OfficeQA Pro62.4%measured
OSWorld-Verified83.9%estimated ± 9.0 pp, low confidence
ParseBench-169.4%measured
Terminal-Bench 4.137.0%measured
Terminal-Bench Hard62.7%estimated ± 2.8 pp, medium confidence
Vals SWE-bench92.0%measured
Terminal-Bench 2.1 (Vals AI)62.9%measured
VulcanBench v378.2%estimated ± 3.9 pp, medium confidence
CWE-Bench-AA55.8%measured
AA Cyber Index49.9%measured
DeepsecBench-AA19.9%measured
AA ITBench51.2%measured
BenchCAD Vision2Code (no tools)0.3%estimated ± 0.1 pp, low confidence
BenchCAD Vision2Code (tools)0.7%estimated ± 0.0 pp, low confidence
Bug Hunt Bench17.7%measured
Chartography (no tools)27.3%estimated ± 6.0 pp, low confidence
Chartography (tools)78.0%measured
CursorBench 4.036.8%measured
CWE-bench v157.4%estimated ± 5.1 pp, medium confidence
DeepSWE63.4%measured
DRACO86.9%estimated ± 0.7 pp, low confidence
FrontierCode 2.027.1%estimated ± 4.7 pp, medium confidence
FrontierPhysics2.5%estimated ± 5.0 pp, low confidence
HLE w/o tools34.1%estimated ± 4.9 pp, high confidence
OfficeQA74.7%estimated ± 0.2 pp, low confidence
OSWorld 2.059.0%estimated ± 5.5 pp, low confidence
OSWorld 2.166.4%estimated ± 6.1 pp, medium confidence
OSWorld 2.1 Offline63.2%estimated ± 7.4 pp, low confidence
PostTrainBench v1.135.6%estimated ± 1.7 pp, low confidence
SpireClimb0.0%estimated ± 9.3 pp, medium confidence
StealthMark47.9%estimated ± 7.7 pp, medium confidence
SWE-bench Multimodal56.6%estimated ± 5.6 pp, low confidence
Terminal-Bench 4.041.3%estimated ± 4.5 pp, high confidence
Terminal-Bench-Science 0.120.8%estimated ± 10.7 pp, low confidence
Toolathlon-Verified78.4%measured
VulcanBench Frontier v485.0%estimated ± 1.8 pp, high confidence
AA AutomationBench54.6%estimated ± 5.5 pp, medium confidence
GDP.pdf19.9%estimated ± 3.1 pp, high confidence
AA Terminal-Bench 4.036.6%estimated ± 6.2 pp, medium confidence
ARC-AGI-30.0%estimated ± 8.6 pp, low confidence
cursorBench4137.9%estimated ± 2.6 pp, medium confidence
CWE-bench44.1%estimated ± 4.3 pp, low confidence
MKA36.2%measured
SkillsBench official leaderboard (BenchFlow)49.4%estimated ± 6.2 pp, medium confidence
CompileBench57.8%estimated ± 10.0 pp, low confidence
FACTS Multimodal36.6%estimated ± 5.6 pp, medium confidence
Hermes Bench68.9%measured
Hermes Index34.9%measured
ParseBench-274.9%estimated ± 4.3 pp, high confidence
CharXiv89.4%measured
CharXiv w/o tools84.5%estimated ± 1.8 pp, medium confidence
ExploitBench16.5%estimated ± 7.2 pp, low confidence
ExploitGym11.9%estimated ± 6.3 pp, medium confidence
AI2D_TEST93.5%estimated ± 0.9 pp, low confidence
CountBench98.3%estimated ± 3.7 pp, low confidence
DynaMath87.9%estimated ± 0.8 pp, low confidence
ERQA70.8%estimated ± 5.6 pp, medium confidence
HMMT Feb 202689.9%estimated ± 1.5 pp, medium confidence
HMMT Nov 202594.8%estimated ± 1.5 pp, medium confidence
MathVision88.6%estimated ± 0.9 pp, high confidence
MCP Atlas82.5%estimated ± 6.0 pp, low confidence
MLVU (M-Avg)88.8%estimated ± 1.5 pp, high confidence
MMAnswerBench88.0%estimated ± 2.4 pp, medium confidence
MMMU-Pro77.9%estimated ± 2.5 pp, medium confidence
NL2Repo56.3%measured
PostTrainBench v140.4%estimated ± 1.3 pp, low confidence
RealWorldQA87.1%estimated ± 1.5 pp, high confidence
ScreenSpot Pro84.5%estimated ± 6.8 pp, medium confidence
Terminal-Bench 2.068.3%estimated ± 5.8 pp, medium confidence
Toolathlon55.9%estimated ± 5.1 pp, low confidence
Video-MME (with subtitle)90.5%estimated ± 2.9 pp, high confidence
VideoMMMU87.1%estimated ± 1.5 pp, medium confidence
We-Math84.6%estimated ± 7.9 pp, low confidence
AA LiveCodeBench99.0%estimated ± 2.6 pp, medium confidence
MedXpertQA (MM)78.8%estimated ± 6.4 pp, low confidence
React Native Evals86.7%estimated ± 4.7 pp, low confidence
PostTrainBench36.8%measured
DAYJOB1.3%estimated ± 3.2 pp, high confidence
DAYJOB Healthcare0.7%estimated ± 4.2 pp, high confidence
Finance Agent v256.1%estimated ± 3.8 pp, medium confidence
FrontierCode 1.1 Extended52.7%estimated ± 2.4 pp, low confidence
HLE-Verified3.2%estimated ± 4.6 pp, low confidence
LABBench280.9%estimated ± 1.8 pp, low confidence
ScarfBench64.0%estimated ± 3.8 pp, medium confidence
BlenderWorld (text-only)71.4%measured
Cookie-Bench76.9%measured
Agents' Last Exam26.3%measured
OpenHarmony Bench v1.057.3%measured
OpenHarmony Bench v1.250.4%measured
SEC-Bench Pro34.4%estimated ± 9.3 pp, medium confidence
BabyVision53.4%measured
MMVU80.5%measured
SWE-Marathon41.1%estimated ± 2.1 pp, low confidence
LifeVal75.2%estimated ± 3.8 pp, medium confidence
MMMU-Pro w/ Python83.0%estimated ± 2.7 pp, medium confidence
ExploitBench Internal Port0.0%estimated ± 1.4 pp, low confidence
MathVision w/ Python95.4%estimated ± 1.0 pp, low confidence
SWE-Atlas Codebase QnA58.2%estimated ± 4.8 pp, high confidence