See key specs and per-benchmark scores for each model/mode. Scroll horizontally for all columns. 当前对比 2 个模型的评测数据与核心参数。

GLM-5.2
智谱AI
Each axis is a category average, normalized to a 100-point radar.
Relative edge: none clear / Relative gap: 综合评估 -14.9
Relative edge: 综合评估 +14.9 / Relative gap: none clear
Method: for each model and benchmark, the chart first averages all scores in the current mode scope instead of taking the best score, then averages those benchmark scores within each category. Only benchmarks with at least two selected models scored are included; missing values are not counted as zero.
Best overall
GLM-5.3 · 56.88
Best single
GLM-5.3 · Terminal-Bench 2.1 88.20
Modality coverage
GLM-5.2 · 1 modalities
Head to head
8
Benchmarks
1
Wins
7
Losses
-5.12
Average diff
Compare benchmark results across thinking modes and tool usage.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
Complete scores for each model/mode across selected benchmarks.
8 benchmarks with comparable scores. Each model shows its best score; mode label is displayed below.
| Benchmark | GLM-5.2 | GLM-5.3 |
|---|---|---|
HLE 综合评估 | 54.70Thinking Enabled | Tools | 62.50Thinking Level · High | Tools |
DeepSWE 编程与软件工程 | 44.00Deep Thinking Mode | Tools | 66.90Thinking Level · High | Tools |
FrontierSWE 编程与软件工程 | 74.40Thinking Level · High | Tools | 78.10Thinking Level · High | Tools |
NL2Repo-Bench 编程与软件工程 | 48.90Thinking Enabled | Tools | 58.00Thinking Level · High | Tools |
PostTrain Bench 编程与软件工程 | 34.30Thinking Level · High | Tools | 39.80Thinking Level · High | Tools |
Program Bench 编程与软件工程 | 63.70Thinking Enabled | Tools | 19.00Thinking Level · High | Tools |
SWE-Marathon 编程与软件工程 | 13.00Thinking Level · High | Tools | 42.50Thinking Level · High | Tools |
Terminal-Bench 2.1 AI Agent - 工具使用 | 81.00Thinking Level · High | Tools | 88.20Thinking Level · High | Tools |
Side-by-side input/output token pricing
Licensing, MoE architecture, and multi-modality support.
| Features & specs | GLM-5.2智谱AI | GLM-5.3智谱AI |
|---|---|---|
Core specsRelease | 2026-06-13 | 2026-08-14 |
Context length | 1M | 1M |
Parameters | 7533.3 | 7533.3 |
Active parameters | 400 | Not provided |
Max output | 128000 | 128000 |
MoE | Yes | Yes |
LicenseCode Open Source | Closed Source | Not provided |
Weights Open Source | Closed Source | Not provided |
Commercial use | 免费商用授权 | Not provided |
Modality supportText Input/Output | / | / |
ResourcesPaper / report | GLM-5: from Vibe Coding to Agentic Engineering | GLM-5.3: Frontier Coding with Emergent Cyber Capabilities |

GLM-5.3
智谱AI