RankingEffort benchmarks

GLM-5.3 · Max

Compare models

62 measured results · 4 effort settings · 49 benchmark/harness combinations

Model evidence and sources →

Change model

Model

Current configurationSources

Compare with

Choose a second model and its effort to see both profiles and benchmark differences below.

Compare withModel and configuration

Capability profile

One configuration, seven capabilities

Comparison score · 0–100

AGLM-5.3 · Max4 of 7 capabilities supported

Gaps are unknown, not zero. Hollow points are preliminary.

Agentic
A60.4Supported
Hard reasoning
A45.4Supported
Coding
A68.7Supported
Human pref
Unknown
Knowledge
A20.9Supported
Multimodal
Unknown
Long context
A70.8Preliminary

Each profile belongs to the named configuration. Capability scores use a shared panel for all axes, distinct from the overall ranking panel; they are not accuracy percentages or direct head-to-head wins.

Capability score evidence

GLM-5.3 · Max: 58.2 overall comparison score. Meets the overall evidence requirements.

5 areas measured. The available task mix affects this estimate; missing areas remain unknown.

Which sources carry the weight?

Cognition · 17.2% direct comparison weight · independent of the model provider
Datacurve · 16.8% direct comparison weight · independent of the model provider
Artificial Analysis · 16.7% direct comparison weight · independent of the model provider
Epoch AI · 16.6% direct comparison weight · independent of the model provider
Terminal-Bench · 16.5% direct comparison weight · independent of the model provider
Vals AI · 16.2% direct comparison weight · independent of the model provider

Which benchmark families carry the weight?

Terminal-Bench · 19.3% direct comparison weight
FrontierCode · 17.2% direct comparison weight
DeepSWE · 16.8% direct comparison weight
GPQA · 4.4% direct comparison weight
Aime · 3.4% direct comparison weight
Other families (24) · 38.8% direct comparison weight

These shares describe direct comparison weight, not fractions of the final score or confidence. Source balancing does not give every family equal weight or remove correlations between overlapping sources.

DeepSWE v1.1 review: Epoch AI reported grading errors in at least 23 of 113 tasks. These published results remain included. Review their limitations in Epoch's review when interpreting this evidence.

How sensitive is this score?

The audit removes one source at a time and refits against the same overall reference panel. These published results use the matching dataset audited 2026-10-23.

Both ranks use the same surviving configurations for each removal. The cohort starts with 444 configurations supported in the original audit, including separate effort settings. These are diagnostic positions, not the default leaderboard ranks. Eligibility is not rechecked after removal.

What changes when a source is removed?
Removed source	Rank before → after	Rank movement	Cohort remaining / lost support	Score after (change)
Artificial Analysis	57 → 54	Up 3	203 remaining / 241 lost	60.1 (+1.9 points)
ARC Prize	59 → 57	Up 2	444 remaining / 0 lost	62.4 (+4.2 points)
Datacurve	59 → 70	Down 11	444 remaining / 0 lost	46.7 (-11.5 points)
LiveCodeBench	59 → 59	No movement	444 remaining / 0 lost	58.2 (+0.0 points)
Scale	59 → 59	No movement	444 remaining / 0 lost	59.1 (+0.9 points)
Terminal-Bench	59 → 60	Down 1	444 remaining / 0 lost	56.8 (-1.4 points)
Cognition	59 → 46	Up 13	444 remaining / 0 lost	57.9 (-0.3 points)
LiveBench	59 → 58	Up 1	443 remaining / 1 lost	59.5 (+1.3 points)
SWE-rebench	59 → 59	No movement	444 remaining / 0 lost	57.6 (-0.6 points)
Epoch AI	59 → 48	Up 11	431 remaining / 13 lost	64.0 (+5.8 points)
Vals AI	59 → 58	Up 1	443 remaining / 1 lost	57.8 (-0.4 points)
Provider reports · Anthropic	59 → 59	No movement	444 remaining / 0 lost	58.1 (-0.1 points)
Provider reports · Moonshot AI	59 → 59	No movement	444 remaining / 0 lost	57.6 (-0.6 points)
Provider reports · Meta	59 → 59	No movement	444 remaining / 0 lost	58.2 (+0.0 points)
Provider reports · Mistral	59 → 59	No movement	444 remaining / 0 lost	58.2 (+0.0 points)
Provider reports · OpenAI	59 → 59	No movement	444 remaining / 0 lost	58.2 (-0.0 points)
All provider reports	59 → 59	No movement	444 remaining / 0 lost	57.5 (-0.7 points)

Lost support means a configuration no longer has a comparison path to the complete reference panel. Its rank is unavailable. A surviving path alone does not establish leaderboard eligibility.

Overlapping sources may remain correlated even with source balancing. Leave-one-source-out movement is a sensitivity diagnostic, not a spread of source-specific ranks or a confidence measure. Scoring methodology · Dated configuration audit

Every benchmark family

Removed evidence	Recomputed score	Change
deepswe	46.7	-11.5
arc-agi	60.5	+2.3
simpleqa	59.9	+1.7
mmlu-pro	59.4	+1.2
mmmu-pro	59.1	+0.9
terminal-bench	57.3	-0.9
vals-tax-agent	57.3	-0.9
epoch-game-puzzles	57.4	-0.8
vals-legal-research	57.4	-0.8
vals-excel-modeling	58.8	+0.6
arc-agi-3	57.6	-0.6
swe-rebench	57.6	-0.6
hle	58.8	+0.6
gpqa	57.6	-0.6
enterprise-ops	58.7	+0.5
gdp-pdf	58.7	+0.5
frontiermath	57.7	-0.4
omniscience	58.6	+0.4
tau-bench	57.8	-0.4
automationbench	57.8	-0.4
vals-finance-agent	57.9	-0.3
multi-swe-bench	57.9	-0.3
swe-atlas-qna	57.9	-0.3
frontiercode	57.9	-0.3
livebench-data	58.5	+0.3
briefcase	57.9	-0.3
swe-atlas-test-writing	57.9	-0.3
gdpval	57.9	-0.3
livebench-instructions	58.4	+0.2
ifbench	58.0	-0.2
livebench-coding	58.0	-0.2
livebench-language	58.4	+0.2
livebench-reasoning	58.1	-0.1
aa-lcr	58.1	-0.1
aime	58.3	+0.1
harvey	58.3	+0.1
enigma-eval	58.3	+0.1
swe-atlas-refactoring	58.3	+0.1
analyst-agent	58.3	+0.1
terminal-bench-science	58.1	-0.1
itbench	58.1	-0.1
critpt	58.2	+0.1
apex-agents	58.2	+0.0
vibe-code	58.1	-0.0
vals-code-migration	58.1	-0.0
livebench-math	58.1	-0.0
swe-bench-pro	58.2	+0.0
math500	58.2	+0.0
gmmlu	58.2	-0.0
livecodebench	58.2	+0.0

What evidence is missing from the fit?

Across all collected settings for this model: 92 observations, 46 contributing, 0 matched without graph weight and 46 excluded. Counts do not establish rank eligibility.

4 · Benchmark family or source protocol has not been reviewed
2 · Original SciCode grading defects are under review; retained as supporting evidence, excluded from aggregate and capability scores. Corrected SciCode-Verified requires a separate protocol review. Sources: https://arxiv.org/abs/2608.04975 and https://artificialanalysis.ai/evaluations/scicode
3 · Source labels low/high/max; Z.ai API effort labels may be metadata only. Bare-bones API column, not ZCode.
6 · Evaluator has not been reviewed for independent admission
1 · Effort is not identified
8 · No reviewed effort for this catalog observation; any separately collected effort measurement is counted separately
1 · No matched, identified effort unit; duplicates and unnamed comparators do not add evidence
13 · Supporting evidence outside the reviewed capability core
1 · The GLM card attributes this evaluation to Proximal; it is not a new Z.ai comparison. Historical provider claim; secondary and superseded for current FrontierSWE comparisons by the independent Proximal FrontierSWE v2 mean@5 board at https://www.frontierswe.com/. Version-unspecified claims are not v2 results. Excluded from weighted Overall.
1 · The source replaces affected anti-cheat checks with LLM inspection and changes task images; the modified protocol requires separate review.
1 · The source changes anti-cheat checks and substitutes zero-shot baseline scores for failed runs; the modified protocol requires separate review.
4 · Narrow specialist biological, medical or security evaluation; preserved as raw evidence outside general capability scoring.
1 · The GLM card attributes this evaluation to Artificial Analysis; it is not a new Z.ai comparison.

Inspect every observation and exclusion →

Benchmark results

These are published benchmark scores. Supported effort estimates also appear on the Capability leaderboard. Results from different harnesses occupy different rows. Missing cells stay unknown; no scores are copied between effort levels.

Benchmark / source	Low	High	Max	Unspecified
AA-AnalystAgentArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of tasks solved on all 5 of 5 attempts · 80 tasks, 5 attempts per task · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	38.75% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
AA-Briefcase EloArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details AA-Briefcase v1.1 is an agentic knowledge work benchmark developed by Artificial Analysis. AA-Briefcase Elo is a combined metric that aggregates rubric pass rate, analytical quality Elo and presentation Elo · Higher is better · Evaluation results measured independently by Artificial Analysis Unit: elo. Source collected 2026-10-18.	1,186.83 Elo Reported settings GLM-5.3 (Low); Model source: https://artificialanalysis.ai/models/glm-5-3-low	Not reported	1,506.86 Elo Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
AA-LCR v1.1Artificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	72.67% Reported settings GLM-5.3 (Low); Model source: https://artificialanalysis.ai/models/glm-5-3-low	Not reported	79.67% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
AutomationBench-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of task objectives completed with no guardrail violations · Higher is better · Benchmark developed by Zapier · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	51.98% Reported settings GLM-5.3 (Low); Model source: https://artificialanalysis.ai/models/glm-5-3-low	Not reported	62.2% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
CritPtArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark developed by Argonne and UIUC, with contributions from 60+ researchers globally. CritPt is under review by its authors following external feedback · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	14.57% Reported settings GLM-5.3 (Low); Model source: https://artificialanalysis.ai/models/glm-5-3-low	Not reported	19.14% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
EnterpriseOps-Gym-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Overall task success rate across all eight enterprise domains (oracle tool mode) · Benchmark developed by ServiceNow Research · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	36.44% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
GDP.pdf: All-pass RateArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of attempts where every atomic criterion passed · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	8.8% Reported settings GLM-5.3 (Low); Model source: https://artificialanalysis.ai/models/glm-5-3-low	Not reported	11.2% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
GDPval-AA v2.1 LeaderboardArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Elo rating for performance on real-world work tasks · Anchored to DeepSeek V4.1 Flash (max) at 1600 · Higher is better · Evaluation results measured independently by Artificial Analysis Unit: elo. Source collected 2026-10-18.	1,322.45 Elo Reported settings GLM-5.3 (Low); Model source: https://artificialanalysis.ai/models/glm-5-3-low	Not reported	1,650.31 Elo Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
GPQA DiamondArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	91.72% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
Harvey LAB-AA v1.1: Hallucination-Gated All-Pass RateArtificial Analysis · Stirrup; Harvey private dataset v1.1.0; 120 tasks across 24 practice areas; up to 200 turns; internet disabled; GPT-6 Sol, Grok 4.7 and Claude Opus 5.5 rubric judges; GPT-6 Sol High two-pass hallucination check; hallucination-gated all-pass averaged across judges Benchmark details Average over tasks of the share of the judge panel finding every rubric criterion passed, with a task zeroed on any material hallucination · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	0.28% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
Humanity's Last ExamArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	36.56% Reported settings GLM-5.3 (Low); Model source: https://artificialanalysis.ai/models/glm-5-3-low	Not reported	42.26% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
ITBench-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Combined public and private Site Reliability Engineering (SRE) tasks · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	46.08% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
MLCR-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of tasks judged accurate, complete, and concise · Expert + compound (hard) sets · Benchmark developed by Wisedocs · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	48.33% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
AA-Omniscience IndexArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details AA-Omniscience Index (higher is better) measures knowledge reliability and hallucination. It rewards correct answers, penalizes hallucinations, and has no penalty for refusing to answer. Scores range from -100 to 100, where 0 means as many correct as incorrect answers, and negative scores mean more incorrect than correct. · Evaluation results measured independently by Artificial Analysis Unit: index. Source collected 2026-10-18.	-8.4 Reported settings GLM-5.3 (Low); Model source: https://artificialanalysis.ai/models/glm-5-3-low	Not reported	14.3 Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
AA-Omniscience AccuracyArtificial Analysis · 6000 questions; 1 attempt; correct-answer share; GPT-5.6 Luna medium grader Benchmark details AA-Omniscience Accuracy (higher is better) measures the proportion of correctly answered questions out of all questions, regardless of whether the model chooses to answer · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	33.92% Reported settings GLM-5.3 (Low); Model source: https://artificialanalysis.ai/models/glm-5-3-low	Not reported	33.85% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
SciCodeArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark curated by scientists across 16 disciplines. SciCode is under review following an independent audit of the dataset · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	42.01% Reported settings GLM-5.3 (Low); Model source: https://artificialanalysis.ai/models/glm-5-3-low	Not reported	59.03% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
𝜏³-BankingArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark developed by Sierra Research · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	50.31% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
Terminal-Bench-Science 0.1Artificial Analysis · mini-swe-agent; Terminal-Bench-Science 0.1.0; 70 tasks; pass@1 averaged over 3 repeats Benchmark details Share of tasks passed (pass@1), averaged over three repeats · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	9.52% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
Terminal-Bench v2.1Artificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark developed by the Laude Institute and open-source contributors · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	83.9% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
Terminal-Bench v4.0Artificial Analysis · mini-SWE-agent v2.4.6; 66 tasks; pass@1 averaged over 3 repeats; 500 steps; task timeout up to 8 hours Benchmark details Benchmark developed by the Laude Institute, Stanford, and open-source contributors · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	34.85% Reported settings GLM-5.3 (Low); Model source: https://artificialanalysis.ai/models/glm-5-3-low	Not reported	41.92% Reported settings GLM-5.3 (Max); Model source: https://artificialanalysis.ai/models/glm-5-3	Not reported
DeepSWE v1.1Datacurve · mini-swe-agent Benchmark details Original public board. Configuration and task coverage may differ; inspect source. Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	69% Reported settings glm-5-3; Reported effort: max; 113 unique tasks; 451 attempts; 4 runs	Not reported
VulcanBench v3 · Report 18VulcanBench · VulcanBench bare-bones API loop, deterministic hidden tests in Docker Benchmark details Frozen v3, 23 post-cutoff repository tasks. Published pass@1; repeat counts and time budgets vary by report. Display only pending comparability review. Unit: percent. Source collected 2026-10-18.	78.3% Reported settings GLM 5.3; source label: low; low effort; 23 tasks; evaluation 2026-08-24. Source labels low/high/max; Z.ai API effort labels may be metadata only. Bare-bones API column, not ZCode. Source labels low/high/max; Z.ai API effort labels may be metadata only. Bare-bones API column, not ZCode. Source label needs review	73.9% Reported settings GLM 5.3; source label: high; high effort; 23 tasks; evaluation 2026-08-24. Source labels low/high/max; Z.ai API effort labels may be metadata only. Bare-bones API column, not ZCode. Source labels low/high/max; Z.ai API effort labels may be metadata only. Bare-bones API column, not ZCode. Source label needs review	65.2% Reported settings GLM 5.3; source label: max; max effort; 23 tasks; evaluation 2026-08-24. Source labels low/high/max; Z.ai API effort labels may be metadata only. Bare-bones API column, not ZCode. Source labels low/high/max; Z.ai API effort labels may be metadata only. Bare-bones API column, not ZCode. Source label needs review	Not reported
FrontierCode 1.1 · extended-chiselCognition · chisel Benchmark details FrontierCode 1.1; chisel. Open each result for its exact effort, task subset, metric and protocol details. Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	52.63% Reported settings FrontierCode 1.1; extended; 150 tasks; chisel; source effort max; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0	Not reported
FrontierCode 1.1 · main-chiselCognition · chisel Benchmark details FrontierCode 1.1; chisel. Open each result for its exact effort, task subset, metric and protocol details. Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	40.14% Reported settings FrontierCode 1.1; main; 100 tasks; chisel; source effort max; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0	Not reported
Terminal-Bench 4.0 · claudecodeTerminal-Bench · Claude Code Benchmark details Terminal-Bench 4.0; Claude Code. Open each result for its exact effort, task subset, metric and protocol details. Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	41.8% Reported settings Terminal-Bench 4.0.0; 66 tasks; Claude Code; effort max; reported margin ±3.2 percentage points, confidence level unspecified; release date 2026-08-14; board checked 2026-10-07	Not reported

What these comparisons mean

Each source reports its own effort labels. Matching labels do not establish equal compute budgets or identical fallback behavior. Open Reported settings to inspect the exact model name, agent and task coverage. Multiple results at one level remain visible rather than selecting the best. Source-reported None or Non-reasoning labels are not treated as evidence that an API supports those settings.

These measurements use their own dated source collection. The leaderboard fits identified configurations jointly; it contains no pooled model ratings. Explicit thinking budgets are separate configurations. Read the methodology.

Download all effort benchmark results · 8,916 results across 259 models · Collected 2026-10-18

GLM-5.3 · Max

Compare with

One configuration, seven capabilities

Agentic

Hard reasoning

Coding

Human pref

Knowledge

Multimodal

Long context

Capability score evidence

Benchmark results