Zhipu AI model product
GLM-4.5-Air is a more compact variant of GLM-4.5 designed for efficient Agentic, Reasoning, and Coding (ARC) applications. It features 106 billion total parameters with 12 billion active parameters using MoE architecture. Like GLM-4.5, it is a hybrid reasoning model providing thinking mode for complex reasoning and tool usage, and non-thinking mode for immediate responses. Despite its compact design, GLM-4.5-Air delivers competitive performance with a score of 59.8 across 12 industry-standard benchmarks, ranking 6th overall while maintaining superior efficiency. It supports 128K context length and is released under MIT open-source license allowing commercial use.
Updated Aug 10, 2026. Default version: GLM-4.5-Air
Structured fields from the published default version.
This profile uses the latest version under this unique model that has a calculated LLMBoard score. Arena and price are excluded.
Published benchmark records for the scored version GLM-4.5-Air.
| AA-Index | 0.6 | 2 | 3 | 50.0% | C | |
| BFCL-v3 | 0.8 | 2 | 19 | 94.4% | C | |
| TAU-bench Airline | 0.6 | 3 | 23 | 90.9% | C | |
| MATH-500 | 1.0 | 4 | 32 | 90.3% | C | |
| TAU-bench Retail | 0.8 | 7 | 25 | 75.0% | C | |
| AIME 2024 | 0.9 | 10 | 53 | 82.7% | C | |
| SciCode | 0.4 | 16 | 18 | 11.8% | C | |
| Terminal-Bench | 0.3 | 20 | 25 | 20.8% | C | |
| LiveCodeBench | 0.7 | 21 | 73 | 72.2% | C | |
| MMLU-Pro | 0.8 | 43 | 129 | 67.2% | C | |
| BrowseComp | 0.2 | 57 | 58 | 1.8% | C | |
| Humanity's Last Exam | 0.1 | 81 | 92 | 12.1% | C | |
| SWE-Bench Verified | 0.6 | 81 | 104 | 22.3% | C | |
| GPQA | 0.8 | 106 | 233 | 54.7% | C |
Preference and agent-evaluation signals from published Arena datasets.
| text | japanese | 106 | 1344.0 | 785 | N/A | |
| text factuality | industry business and management and financial operations | 116 | 1395.6 | 751 | N/A | |
| text factuality | industry writing and literature and language | 118 | 1374.9 | 953 | N/A | |
| text factuality | multi turn | 118 | 1384.6 | 702 | N/A | |
| text factuality | hard prompts english | 120 | 1427.3 | 1,130 | N/A | |
| text | french | 121 | 1407.6 | 466 | N/A | |
| text factuality | coding | 122 | 1429.4 | 859 | N/A | |
| text factuality | instruction following | 123 | 1379.0 | 1,149 | N/A | |
| text style control | japanese | 125 | 1327.4 | 785 | N/A | |
| text | chinese | 126 | 1431.2 | 1,614 | N/A | |
| text factuality | longer query | 126 | 1386.9 | 1,040 | N/A | |
| text factuality | non english | 130 | 1382.7 | 3,225 | N/A | |
| text factuality | exclude ties | 131 | 1382.6 | 4,874 | N/A | |
| text factuality | overall | 132 | 1396.1 | 7,198 | N/A | |
| text factuality | english | 132 | 1405.4 | 3,072 | N/A | |
| text factuality | industry software and it services | 133 | 1417.6 | 1,499 | N/A | |
| text | german | 134 | 1370.4 | 672 | N/A | |
| text | polish | 134 | 1367.4 | 1,851 | N/A | |
| text factuality | hard prompts | 134 | 1407.0 | 3,544 | N/A | |
| text | industry life and physical and social science | 136 | 1407.9 | 4,924 | N/A | |
| text | industry medicine and healthcare | 136 | 1403.6 | 1,641 | N/A | |
| text | math | 136 | 1397.4 | 1,535 | N/A | |
| text style control | french | 136 | 1404.6 | 466 | N/A | |
| text | spanish | 138 | 1372.4 | 723 | N/A | |
| text | industry mathematical | 141 | 1393.9 | 1,447 | N/A | |
| text | english | 143 | 1395.0 | 14,806 | N/A | |
| text | overall | 144 | 1382.7 | 31,040 | N/A | |
| text | exclude ties | 144 | 1356.3 | 21,840 | N/A | |
| text | industry legal and government | 145 | 1386.2 | 1,933 | N/A | |
| text | industry business and management and financial operations | 147 | 1371.1 | 5,505 | N/A | |
| text | non english | 147 | 1366.7 | 16,184 | N/A | |
| text | russian | 147 | 1369.3 | 1,482 | N/A | |
| text | industry software and it services | 148 | 1400.5 | 10,645 | N/A | |
| text | expert | 149 | 1368.3 | 1,388 | N/A | |
| text | korean | 149 | 1303.2 | 635 | N/A | |
| text style control | chinese | 149 | 1412.9 | 1,614 | N/A | |
| text | hard prompts | 150 | 1378.2 | 14,862 | N/A | |
| text | hard prompts english | 150 | 1387.0 | 7,644 | N/A | |
| text | industry writing and literature and language | 150 | 1354.0 | 7,011 | N/A | |
| text style control | math | 150 | 1389.3 | 1,535 | N/A | |
| text | coding | 151 | 1396.5 | 6,099 | N/A | |
| text style control | industry mathematical | 152 | 1390.1 | 1,447 | N/A | |
| text | creative writing | 153 | 1341.9 | 4,010 | N/A | |
| text | multi turn | 153 | 1371.9 | 5,120 | N/A | |
| text | industry entertainment and sports and media | 154 | 1338.6 | 5,674 | N/A | |
| text | instruction following | 156 | 1353.8 | 8,107 | N/A | |
| text | longer query | 157 | 1365.1 | 6,613 | N/A | |
| text style control | german | 160 | 1348.4 | 672 | N/A | |
| text style control | spanish | 160 | 1363.3 | 723 | N/A | |
| text style control | industry life and physical and social science | 161 | 1396.9 | 4,924 | N/A | |
| text style control | industry software and it services | 162 | 1414.4 | 10,645 | N/A | |
| text style control | coding | 163 | 1426.5 | 6,099 | N/A | |
| text style control | industry business and management and financial operations | 163 | 1371.9 | 5,505 | N/A | |
| text style control | english | 164 | 1389.1 | 14,806 | N/A | |
| text style control | expert | 164 | 1387.8 | 1,388 | N/A | |
| text style control | industry medicine and healthcare | 164 | 1396.1 | 1,641 | N/A | |
| text style control | longer query | 164 | 1378.9 | 6,613 | N/A | |
| text style control | hard prompts english | 166 | 1402.8 | 7,644 | N/A | |
| text style control | polish | 167 | 1341.2 | 1,851 | N/A | |
| text style control | hard prompts | 168 | 1390.6 | 14,862 | N/A | |
| text style control | overall | 169 | 1372.7 | 31,040 | N/A | |
| text style control | instruction following | 169 | 1362.1 | 8,107 | N/A | |
| text style control | multi turn | 169 | 1369.2 | 5,120 | N/A | |
| text style control | korean | 170 | 1285.1 | 635 | N/A | |
| text style control | russian | 170 | 1356.9 | 1,482 | N/A | |
| text style control | exclude ties | 171 | 1342.9 | 21,840 | N/A | |
| text style control | non english | 172 | 1352.4 | 16,184 | N/A | |
| text style control | industry legal and government | 179 | 1374.3 | 1,933 | N/A | |
| text style control | industry writing and literature and language | 179 | 1341.7 | 7,011 | N/A | |
| text style control | creative writing | 180 | 1326.9 | 4,010 | N/A | |
| text style control | industry entertainment and sports and media | 180 | 1325.0 | 5,674 | N/A |
Official vendor API PAYG pricing is summarized first. The table then lists individual provider offerings without treating their minimum as the official price.
| 302.AI | glm-4.5-air | global | $0.1143 | $0.286 | 131.1K | |
| LLM Gateway | glm-4.5-air | global | $0.13 | $0.85 | 131K | |
| Z.AI | glm-4.5-air | global | $0.20 | $1.1 | 131.1K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
All published versions linked to this unique model. The score columns identify the version used by the current overall ranking.
| GLM-4.5-Air | 55.1 | 106B | 131.1K | 98.3K | Yes | MIT |
A concise description based on the published model registry.
GLM-4.5-Air is a more compact variant of GLM-4.5 designed for efficient Agentic, Reasoning, and Coding (ARC) applications. It features 106 billion total parameters with 12 billion active parameters using MoE architecture. Like GLM-4.5, it is a hybrid reasoning model providing thinking mode for complex reasoning and tool usage, and non-thinking mode for immediate responses. Despite its compact design, GLM-4.5-Air delivers competitive performance with a score of 59.8 across 12 industry-standard benchmarks, ranking 6th overall while maintaining superior efficiency. It supports 128K context length and is released under MIT open-source license allowing commercial use.
Use the benchmark, Arena and pricing sections above as separate evidence. A missing field means the current data snapshot does not support that claim.
Data snapshot: 2026-08-07. Editorial model content is not available in the backend.
Open a comparison with the three ranked models immediately above and below this model.
Recommendations prioritize the same model type and family, then the closest published LLMBoard score.
Common questions about GLM 4.5 Air.
GLM 4.5 Air's default version was released on Jul 28, 2025.
No official standard PAYG price is currently available for GLM 4.5 Air. The lowest tracked third-party offer starts at $0.1143 input and $0.286 output via 302.AI.
GLM 4.5 Air is published under Zhipu AI in the model registry.
The default version has a 131.1K token context window.
Yes. The default version is marked as open weight under MIT.
4 published provider offerings are linked to the default version.
Nearby ranked alternatives include MiniMax M1 80K, MiMo V2 Flash, o3.