GLM-4.5-Air is a 106B total / 12B active parameter model designed to unify frontier reasoning, coding, and agentic capabilities. On the SWE-bench Verified benchmark, it delivers the best performance at its scale with a competitive performance-to-cost ratio.
Added Apr 15, 2025
Model weightsContext Window
128.0K
Max Output
98.3K
Avg output tokens (7d)
182 tokens
Input Price (Auto)
$0.12/1M
Output Price (Auto)
$0.80/1M
Cache Read (Auto)
$0.060/1M
Capabilities
Benchmarks
Benchmarks
Performance metrics and benchmarks
Sourced from Artificial Analysis.
Intelligence Index
16.7
Reasoning
GPQA Diamond
Graduate-level scientific reasoning
73.3%
Better than 56% of models compared
HLE
Humanity's Last Exam
7.0%
Better than 48% of models compared
IFBench
Instruction-following benchmark
37.6%
Better than 32% of models compared
T²-Bench Telecom
Conversational AI agents in dual-control scenarios
46.5%
Better than 50% of models compared
AA-LCR
Long context reasoning evaluation
45.7%
Better than 51% of models compared
CritPt
Research-level physics reasoning
0.0%
Coding
SciCode
Python programming for scientific computing
30.6%
Better than 43% of models compared
Terminal-Bench Hard
Agentic coding and terminal use
20.5%
Better than 60% of models compared
LiveCodeBench
Contamination-free coding benchmark
68.4%
Better than 76% of models compared
Math
AIME 2025
American Invitational Mathematics Examination 2025
80.7%
Better than 77% of models compared
AIME
American Invitational Mathematics Examination
67.3%
Better than 77% of models compared
Math-500
Diverse mathematical problem solving benchmark
96.5%
Better than 84% of models compared
Knowledge
MMLU-Pro
Professional and academic subject knowledge
81.5%
Better than 75% of models compared
AA-Omniscience Accuracy
Proportion of correctly answered questions
16.3%
AA-Omniscience Hallucination Rate
Rate of incorrect answers among non-correct responses
92.9%
Last updated Aug 18, 2026
Artificial AnalysisProviders
Choose explicit providers for this model. Auto routing remains available as the default option.
Loading provider options…
Related text models
Compare GLM 4.5 Air with similar models from the same provider or model family.
GLM 4.5 Air (Thinking)
zai-org/GLM-4.5-Air:thinkingGLM-4.5-Air with thinking mode enabled for enhanced reasoning capabilities. Shows step-by-step thought process.
GLM 5.3
zai-org/glm-5.3GLM-5.3 for long-horizon autonomous coding and engineering workflows. This variant defaults to the model's low reasoning tier for faster responses.
GLM 5.3 Thinking
zai-org/glm-5.3:thinkingGLM-5.3 with higher reasoning enabled for harder long-horizon coding, autonomous agent workflows, and complex engineering tasks.
GLM 5.2
zai-org/glm-5.2GLM-5.2 is Z.AI's flagship model for long-horizon autonomous coding and engineering workflows. It is built to plan, execute, iterate, and optimize complex development tasks over extended runs. This variant keeps thinking disabled for faster direct responses.
GLM 5.2 Thinking
zai-org/glm-5.2:thinkingGLM-5.2 with thinking enabled for harder long-horizon coding, autonomous agent workflows, complex engineering optimization, and real-world development tasks.
GLM Latest
zai-org/glm-latestCompatibility alias that routes to the newest thinking GLM model. Currently routes to GLM 5.2 Thinking.