Skip to main content

Model profile

GLM-5.2

Z.AICurrentReleased Jun 16, 2026
Data verified
Overall Score
63.96Public #37 of 200
Arena Elo
1470
Eligible category ranks
3of 8
Price (1M tokens)
$1.4 in / $4.4 out
API pricing
Speed
Not listed
Context
1M

Evidence coverage

43 of 323 tracked benchmarks are published. 18 are verified and 25 provisional. 7 of 8 categories are measured.

Updated July 23, 2026Methodology
Published / tracked
43 / 323
Verified
18
Provisional
25
Categories with evidence
7 / 8

Evidence by category

  • Agentic17 benchmarks
    Mixed evidence
  • Coding7 benchmarks
    Mixed evidence
  • Reasoning2 benchmarks
    Mixed evidence
  • Knowledge11 benchmarks
    Mixed evidence
  • Math4 benchmarks
    Verified
  • Multilingual0 benchmarks
    Not measured
  • Multimodal1 benchmark
    Reported
  • Inst. Following1 benchmark
    Reported
Open WeightSelf-hostReasoning
Confidence:
Medium
flagship

GLM-5.2 ranks #37 out of 200 models on the public leaderboard with an overall score of 63.96/100. It does not yet have enough sourced coverage for BenchLM's verified leaderboard. While not a frontier model, it offers specific advantages depending on the use case.

GLM-5.2 is a open weight model with a 1M token context window. It uses explicit chain-of-thought reasoning, which typically improves performance on math and complex reasoning tasks at the cost of higher latency and token usage.

GLM-5.2 sits inside the GLM-5 family alongside GLM-5, GLM-5.1, GLM-5 (Reasoning), GLM-5-Turbo, GLM-5V-Turbo. BenchLM links it directly to GLM-5.1 as the earlier related model in that lineage. This profile currently has 43 of 323 tracked benchmarks. BenchLM only exposes non-generated benchmark rows publicly, so missing categories stay blank until a sourced evaluation is available.

Its strongest category is Knowledge (#6), while its weakest is Agentic (#24). This performance profile makes it particularly effective for knowledge-intensive tasks like research, analysis, and factual Q&A.

Peer position

Exact provisional scores and ranks for the closest listed peers. A score can appear before a model clears the evidence threshold for a rank, so equal scores can have different rank states.

Range 63.564.75

  1. Gemini 3.5 Flash
    Google
    #3364.75
    Gemini 3.5 Flash is #33 with a score of 64.75.
    Compare
  2. Claude Opus 4.5
    Anthropic
    #3464.22
    Claude Opus 4.5 is #34 with a score of 64.22.
    Compare
  3. Claude Opus 4.6 (Adaptive)
    Anthropic
    #3564.18
    Claude Opus 4.6 (Adaptive) is #35 with a score of 64.18.
    Compare
  4. MiniMax M2.7
    MiniMax
    #3664.11
    MiniMax M2.7 is #36 with a score of 64.11.
    Compare
  5. GLM-5.2Current model
    Z.AI
    #3763.96
    GLM-5.2 is #37 with a score of 63.96.
  6. GPT-5.5 Pro
    OpenAI
    #3863.69
    GPT-5.5 Pro is #38 with a score of 63.69.
    Compare
  7. GLM-5V-Turbo
    Z.AI
    #3963.5
    GLM-5V-Turbo is #39 with a score of 63.5.
    Compare

Category percentile

More

Relative position among models eligible for each sourced category. A higher percentile means a stronger position within that category's ranked cohort; 100 is highest.

  1. Knowledge90%
    Eligible cohort rank #6 of 52Category score 87.0
  2. Coding91%
    Eligible cohort rank #12 of 122Category score 65.1
  3. Agentic81%
    Eligible cohort rank #24 of 119Category score 54.6

Category evidence

Scores and ranks appear only where this model has published benchmark evidence. Categories without displayable source records remain not measured.

Category scores, ranks, weighting, benchmark coverage, and evidence status
CategoryScore
AgenticRank #24 of 119Percentile 81stWeight 22%17 benchmarksMixed sources54.6
CodingRank #12 of 122Percentile 91stWeight 20%7 benchmarksMixed sources65.1
ReasoningWeight 17%2 benchmarksMixed sourcesScore pending
KnowledgeRank #6 of 52Percentile 90thWeight 12%11 benchmarksMixed sources87.0
MathRank Not rankedWeight 5%4 benchmarksVerified81.4
MultilingualWeight 7%0 benchmarksNot measuredNot measured
MultimodalWeight 12%1 benchmarkReportedScore pending
Inst. FollowingWeight 5%1 benchmarkReportedScore pending

Chatbot Arena performance

Scroll horizontally to inspect confidence intervals and vote counts.

Chatbot Arena Elo, confidence interval, and vote count by evaluation view
ViewEloConfidence intervalVotes
Text Overall1470±5.917,103
Coding1511±9.34,838
Math1475±20.2829
Instruction Following1464±8.55,859
Creative Writing1447±11.53,049
Multi-turn1471±11.72,814
Hard Prompts1490±7.011,254
Hard Prompts (English)1490±8.85,407
Longer Query1481±8.07,693

Benchmark Details

Rows below have a displayable published verification record. Each source link and provenance note remains in the page HTML while its category is closed. Source-unverified manual rows and generated rows stay hidden.

Agentic17 benchmarks
Terminal-Bench 2.0Provider exact
81%Weighted 38%
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GLM-5.2 at 81.0 on Terminal Bench 2.1 using Terminus-2.
MCP AtlasProvider exact
76.8%Display only
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GLM-5.2 at 76.8 on MCP-Atlas Public Set.
ToolathlonProvider exact
48.2%Display only
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GLM-5.2 at 48.2 on Tool-Decathlon. BenchLM stores this on the existing Toolathlon display key.
AA Agentic IndexReported

Artificial Analysis Agentic Index

43.1%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
τ²-bench resultsReported

τ²-Bench Tool-Agent-User Evaluation

99.1%Display only
Source: Artificial Analysis: tau2-bench leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
GDPval-AAReported

GDPval-AA normalized

50.7%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
GDPval-AAReported
1514Display only
Source: Artificial Analysis: gdpval-aa leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
APEX-Agents-AAReported
33.7%Display only
Source: Artificial Analysis: apex-agents-aa leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
ResearchClawBenchBenchmark exact
20.7%Display only
Source: ResearchClawBench leaderboardProvenance: ResearchClawBench reports this model as ResearchHarness (GLM-5.2) in the official Pass@1 leaderboard. BenchLM stores the one-decimal RADS average on the local ResearchClawBench display key and excludes it from weighted rankings.
AA BriefcaseReported

Artificial Analysis Briefcase

1260Display only
Source: Artificial Analysis: aa-briefcase leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
AA AutomationBenchReported

Artificial Analysis AutomationBench

27.8%Display only
Source: Artificial Analysis: automationbench-aa leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
AA EnterpriseOps-GymReported

Artificial Analysis EnterpriseOps-Gym

42.7%Display only
Source: Artificial Analysis: enterprise-ops-gym-aa leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
AA Harvey LABReported

Artificial Analysis Harvey LAB-AA

91.0%Display only
Source: Artificial Analysis: harvey-lab-aa leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
AA ITBenchReported

Artificial Analysis ITBench-AA

42.7%Display only
Source: Artificial Analysis: itbench-aa leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
AA Tau3 BankingReported

Artificial Analysis Tau3-Banking

26.8%Display only
Source: Artificial Analysis: tau3-banking leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
terminalBenchHardReported
50.8%Display only
Source: Artificial Analysis: terminalbench-hard leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
aaTerminalBench21Reported
77.9%Display only
Source: Artificial Analysis: terminalbench-v2-1 leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
Coding7 benchmarks
SWE-bench ProProvider exact
62.1%Weighted 10%
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GLM-5.2 at 62.1 on SWE-bench Pro.
NL2RepoProvider exact
48.9%Display only
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GLM-5.2 at 48.9 on NL2Repo.
Terminal-Bench 2.0Provider exact
81.0%Display only
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GLM-5.2 at 81.0 on Terminal Bench 2.1 using Terminus-2. BenchLM stores it in the existing Terminal-Bench 2 slot.
ProgramBenchProvider exact

ProgramBench: Can Language Models Rebuild Programs From Scratch?

63.7%Display only
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GLM-5.2 at 63.7 on ProgramBench. BenchLM stores this display-only provider-reported row in the existing ProgramBench key.
cursorBench32Benchmark exact
55.0%Display only
Source: Cursor evals: CursorBench 3.2Provenance: Cursor reports GLM 5.2 Max at this exact CursorBench 3.2 score on its public evals page. BenchLM stores it on the GLM-5.2 row as a display-only coding-agent benchmark.
AA Coding IndexReported

Artificial Analysis Coding Index

68.8%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-SciCodeReported

Artificial Analysis SciCode

50.5%Display only
Source: Artificial Analysis: scicode leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
Reasoning2 benchmarks
CritPtProvider exact

Critical Physics Tasks

20.9%Display only
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GLM-5.2 at 20.9 on CritPt.
AA-LCRReported

Artificial Analysis Long Context Reasoning

71.3%Display only
Source: Artificial Analysis: artificial-analysis-long-context-reasoning leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
Knowledge11 benchmarks
HLEProvider exact

Humanity's Last Exam

54.7%Weighted 45%
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GLM-5.2 at 54.7 on HLE with tools. BenchLM follows the existing GLM-5.1 mapping and stores the tool-assisted result on the weighted HLE lane.
GPQAProvider exact

Graduate-Level Google-Proof Q&A

91.2%Weighted 7%
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GPQA-Diamond at 91.2. BenchLM stores that exact value on the weighted GPQA lane and the display GPQA-Diamond lane for consistency with other provider launch tables.
GPQA-DProvider exact

GPQA Diamond

91.2%Display only
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GPQA-Diamond at 91.2.
HLE w/o toolsProvider exact

Humanity's Last Exam without tools

40.5%Display only
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GLM-5.2 at 40.5 on HLE without tools.
Artificial Analysis Intelligence IndexReported
51.1%Display only
Source: Artificial Analysis: artificial-analysis-intelligence-index leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
AA-GPQA DiamondReported

Artificial Analysis GPQA Diamond

89.5%Display only
Source: Artificial Analysis: gpqa-diamond leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
AA-HLEReported

Artificial Analysis Humanity's Last Exam

40.1%Display only
Source: Artificial Analysis: humanitys-last-exam leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
AA-Omniscience IndexReported

Artificial Analysis Omniscience Index

4.0%Display only
Source: Artificial Analysis: omniscience leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
AA-Omniscience AccuracyReported

Artificial Analysis Omniscience Accuracy

25.1%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA-Omniscience Hallucination RateReported

Artificial Analysis Omniscience Hallucination Rate

28.1%Display only
Source: Artificial Analysis model benchmarksProvenance: Display-only row synced from the current Artificial Analysis model payload. It is excluded from BenchLM weighted scoring.
AA Openness IndexReported

Artificial Analysis Openness Index

44.4%Display only
Source: Artificial Analysis: artificial-analysis-openness-index leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.
Math4 benchmarks
AIME26Provider exact

AIME 2026

99.2%Weighted 25%
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GLM-5.2 at 99.2 on AIME 2026.
HMMT Feb 2026Provider exact

Harvard-MIT Mathematics Tournament February 2026

92.5%Weighted 25%
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GLM-5.2 at 92.5 on HMMT Feb. 2026.
HMMT Nov 2025Provider exact

Harvard-MIT Mathematics Tournament November 2025

94.4%Display only
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GLM-5.2 at 94.4 on HMMT Nov. 2025.
MMAnswerBenchProvider exact
91.0%Display only
Source: Z.AI GLM-5.2 model cardProvenance: Z.AI reports GLM-5.2 at 91.0 on IMOAnswerBench. BenchLM stores it on the existing mmAnswerBench key to match the GLM-5.1 precedent.
Multimodal1 benchmark
Design Arena WebsiteReported

Design Arena Website Elo

1340Display only
Source: OpenRouter model benchmarksProvenance: Display-only Design Arena Website Elo synced from OpenRouter model benchmark metadata. It is excluded from BenchLM weighted scoring.
Inst. Following1 benchmark
AA-IFBenchReported

Artificial Analysis IFBench

73.3%Display only
Source: Artificial Analysis: ifbench leaderboardProvenance: Display-only row synced from the current Artificial Analysis evaluation leaderboard. It is excluded from BenchLM weighted scoring.

Frequently Asked Questions

How does GLM-5.2 perform overall in AI benchmarks?

GLM-5.2 currently ranks #37 out of 200 models on BenchLM's provisional leaderboard with an overall score of 63.96. It is created by Z.AI. Its published context window is 1M.

Is GLM-5.2 good for knowledge and understanding?

GLM-5.2 ranks #6 out of 52 models in knowledge and understanding benchmarks with an average score of 87. It is among the top performers in this category.

Is GLM-5.2 good for coding and programming?

GLM-5.2 ranks #12 out of 122 models in coding and programming benchmarks with an average score of 65.1. There are stronger options in this category.

Is GLM-5.2 good for mathematics?

GLM-5.2 has visible benchmark coverage in mathematics, but BenchLM does not currently assign it a global category rank there.

Is GLM-5.2 good for reasoning and logic?

GLM-5.2 has visible benchmark coverage in reasoning and logic, but BenchLM does not currently assign it a global category rank there.

Is GLM-5.2 good for agentic tool use and computer tasks?

GLM-5.2 ranks #24 out of 119 models in agentic tool use and computer tasks benchmarks with an average score of 54.6. There are stronger options in this category.

Is GLM-5.2 good for multimodal and grounded tasks?

GLM-5.2 has visible benchmark coverage in multimodal and grounded tasks, but BenchLM does not currently assign it a global category rank there.

Is GLM-5.2 good for instruction following?

GLM-5.2 has visible benchmark coverage in instruction following, but BenchLM does not currently assign it a global category rank there.

Is GLM-5.2 open source?

Yes, GLM-5.2 is an open weight model created by Z.AI, meaning it can be downloaded and run locally or fine-tuned for specific use cases.

Which sibling models are related to GLM-5.2?

GLM-5.2 belongs to the GLM-5 family. Related variants on BenchLM include GLM-5, GLM-5.1, GLM-5 (Reasoning), GLM-5-Turbo, GLM-5V-Turbo.

Does GLM-5.2 have full benchmark coverage on BenchLM?

Not yet. GLM-5.2 currently has 43 published benchmark scores out of the 323 benchmarks BenchLM tracks. BenchLM only exposes non-generated public benchmark rows, so missing categories stay blank until a sourced evaluation is available.

What is the context window size of GLM-5.2?

GLM-5.2 has a published context window of 1M, which determines how much text it can process in a single interaction.

Last updated: July 23, 2026 · Runtime metrics stay blank until BenchLM has a sourced snapshot.

Choose with this week’s evidence

Join 2,000+ readers for ranking moves, new releases, pricing changes, and the evidence behind them.

Free. One email per week.