GLM-5.3
- Best result
- 91.54onGPQA diamondunverified
- Standing
- #42 of 183 on GPQA diamond — top quarter of the field
- Leads
- 0 of 16 benchmarks
- Evidence
- 0 of 16 scores independently reproduced
- third-party
- unverified
Each bar is drawn on its own benchmark's scale — the ceiling that benchmark states where it has one (of 100), otherwise the best score recorded on it (rel. 2103). A benchmark that states neither draws no bar.
| Benchmark | Score | Evidence status | Released |
|---|---|---|---|
| GPQA diamond | 91.54accuracy (%) | unverifiedT2 | 2026-08-14 |
| OTIS Mock AIME 2024-2025 | 91.1accuracy (%) | unverifiedT2 | 2026-08-14 |
| DTBench | 79.45accuracy (%) | third-partyT1 | 2026-08-14https://conceptualreasoning.ai/dtbench |
| WeirdML | 75.4accuracy (%) | third-partyT1 | 2026-08-14https://htihle.github.io/weirdml.html |
| FrontierMath-Tiers-1-3-v2-Private | 72.39accuracy (%) | unverifiedT2 | 2026-08-14 |
| LMCA | 65.31accuracy (%) | third-partyT1 | 2026-08-14https://conceptualreasoning.ai/lmca |
| SimpleBench | 59.44accuracy (%) | third-partyT1 | 2026-08-14https://simple-bench.com/ |
| APEX-Agents | 56.6accuracy (%) | unverifiedT2 | 2026-08-14 |
| ProofBench | 49accuracy (%) | third-partyT1 | 2026-08-14https://www.vals.ai/benchmarks/proof_bench |
| SimpleQA Verified | 41accuracy (%) | unverifiedT2 | 2026-08-14 |
| FrontierCode | 40.14accuracy (%) | third-partyT1 | 2026-08-14https://cognition.com/frontiercode |
| PostTrainBench | 34.52accuracy (%) | unverifiedT2 | 2026-08-14 |
| Chess Puzzles | 33.71accuracy (%) | unverifiedT2 | 2026-08-14 |
| FrontierSWE | 30.18accuracy (%) | third-partyT1 | 2026-08-14https://www.frontierswe.com/ |
| FrontierMath-Tier-4-v2-Private | 29.27accuracy (%) | unverifiedT2 | 2026-08-14 |
| Mystery Game Puzzles | 26.2accuracy (%) | unverifiedT2 | 2026-08-14 |
Claims drawn from cited facts, not live model generation.
This model was developed by Zhipu AI and released in 2026.
2 cited facts
This model is scored on 16 tracked benchmarks but holds no current top score. Its record is currently unverified: neither confirmed nor disputed by outside evaluation.
3 cited facts
Across its benchmarks, the model trails the top reported score by an average of 31.72 points under the disclosed harness — a wide gap that reads as well behind the leaders rather than competitive or at the frontier, and a score gap only, not a capability claim. It holds the top score on 0 benchmarks, so there is no genuine category leadership yet on any measured task.
2 cited facts
1 facts cross-checked across data sources: 1 single-source.
Source facts, citations, and refresh stamp for this record.
Sources: epoch_benchmarks
GLM-5.3 is an AI model developed by Zhipu AI, released 2026-08-14. Its benchmark record below tags every score as independently reproduced, third-party, self-reported, or unverified.
GLM-5.3 has recorded scores on 16 benchmarks, each shown with its evidence status.
GLM-5.3 has recorded scores on 16 benchmarks — ProofBench, GPQA diamond, OTIS Mock AIME 2024-2025, DTBench, WeirdML, LMCA, and 10 more. The full table above shows each score with its evidence status.
0 of 16 tracked claims (0%) are independently reproduced; the rest are third-party, self-reported, unverified, or contradicted by an independent measurement.
GLM-5.3 has 16 tracked claims: 9 unverified.
Plain text
tensor.news. "GLM-5.3 — benchmark record". Benchmark record with evidence status for every score Updated 2026-11-18. https://tensor.news/models/glm-5-3
BibTeX
@misc{tensornews_model_glm-5-3-benchmark-record,
title = {{GLM-5.3 — benchmark record}},
author = {{tensor.news}},
year = {2026},
month = {11},
howpublished = {\url{https://tensor.news/models/glm-5-3}},
note = {Benchmark record with evidence status for every score},
urldate = {2026-11-18},
}