The Model Index
Up to four models

GLM-5.2 vs GLM-5.1

On the index (GPT-4 = 100), GLM-5.2 scores 154 and GLM-5.1 151. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.

Next to GLM-5.2
Presets
GLM-5.2
154138–169
WeightsOpen
ReleasedJun 2026
Price in / out$0.0070 / $6.8
Context1.05M
Core tests taken6 of 9
GLM-5.1
151134–167
WeightsOpen
ReleasedApr 2026
Price in / out$1.4 / $4.4
Context205K
Core tests taken5 of 9

GLM-5.2 and GLM-5.1 are too close to call (3 points apart, within the uncertainty), ahead on 5 of 5 shared core tests, at a similar price.

Test by testthird-party results

GLM-5.2GLM-5.1
0%25%50%75%100%REASONING & KNOWLEDGEGPQA Diamond (Epoch AI)GPQA Diamond (Epoch AI)GPQA Diamond (Vals AI)GPQA Diamond (Vals AI)MMLU ProMMLU ProSimpleBenchSimpleBenchSimpleQA VerifiedSimpleQA VerifiedMATHSFrontierMath (tiers 1-3)FrontierMath (tiers 1-3)AIME-style maths (OTIS mock)AIME-style maths (OTIS mock)Chess puzzlesChess puzzlesCODINGSWE-bench VerifiedSWE-bench VerifiedCode MigrationCode MigrationLiveCodeBenchLiveCodeBenchProgramBenchProgramBenchSWE-benchSWE-benchVibe Code Bench v1.1Vibe Code Bench v1.1WeirdMLWeirdMLAGENTSAPEX-AgentsAPEX-AgentsPROFESSIONAL WORKFinance Agent (v2)Finance Agent (v2)CorpFin v2CorpFin v2LegalBenchLegalBenchLegal Research BenchLegal Research BenchMedCodeMedCodeMedScribeMedScribeTaxEval v2TaxEval v2
Every result as a table
TestGLM-5.2GLM-5.1
Reasoning & knowledge
GPQA Diamond (Epoch AI)91.9%89.9%
GPQA Diamond (Vals AI)85.6%84.5%
MMLU Pro86.7%86.9%
SimpleBench58.8%55.1%
SimpleQA Verified34.2%34.0%
Maths
FrontierMath (tiers 1-3)59.2%36.8%
AIME-style maths (OTIS mock)86.4%93.3%
Chess puzzles22.0%20.0%
Coding
SWE-bench Verified78.7%74.2%
Code Migration37.9%25.8%
LiveCodeBench69.5%81.4%
ProgramBench0.5%0.0%
SWE-bench82.8%76.4%
Vibe Code Bench v1.164.0%31.5%
WeirdML70.1%57.1%
Agents
APEX-Agents45.2%40.9%
Professional work
Finance Agent (v2)49.7%44.8%
CorpFin v266.1%64.5%
LegalBench84.1%84.4%
Legal Research Bench31.3%27.9%
MedCode40.8%41.6%
MedScribe83.5%72.3%
TaxEval v273.3%71.2%

Tests at least two of these models have taken. Bold rows are the core tests.

This page as a chartDownload PNG