Z.ai (Zhipu), open weights
GLM-4.5-Air
GLM-4.5-Air by Z.ai (Zhipu) ranks 177th of 354 ranked models on the Noometry Index as of October 2026, with a score of 38.9. Its strongest category is long context, where it ranks 135th. API pricing starts at $0.20 per million input tokens and $1.10 per million output tokens, with a 131K-token context window.
Last verified
Specifications
- Noometry rank
- #177 of 354
- Index score
- 38.9
- Evidence
- Confirmed 27 results
- Provider
- Z.ai (Zhipu)
- Released
- July 20, 2025
- Weights
- Open weights
- Reasoning
- Yes
- Context window
- 131K
- Max output
- 98K
- Input price
- $0.20 / M
- Output price
- $1.10 / M
- Blended price
- $0.43 / M
- Output speed
- 160 tokens/s Kagi
- Value
- #70 of 219
- Knowledge cutoff
- April 2025
- Input
- text
- Hugging Face
- zai-org/GLM-4.5-Air
Category scores
Each category score combines every public result we have in that category.
- Coding 33.3
- Reasoning 24.1
- Math 36.2
- Knowledge 35.0
- Multilingual 49.1
- Instruction Following 69.6
- Long Context 41.6
- Writing & Preference 55.9
| Category | Score | Rank | Results |
|---|---|---|---|
| Coding | 33.3 | #259 | 2 |
| Reasoning | 24.1 | #166 | 2 |
| Math | 36.2 | #170 | 2 |
| Knowledge | 35.0 | #191 | 5 |
| Multilingual | 49.1 | #135 | 1 |
| Instruction Following | 69.6 | #171 | 2 |
| Long Context | 41.6 | #135 | 1 |
| Writing & Preference | 55.9 | #139 | 4 |
Strengths and weaknesses
Categories where GLM-4.5-Air places highest and lowest among the models ranked in each, with its score against that category's median.
Strongest categories
| Category | Score | vs median | Rank |
|---|---|---|---|
| Writing & Preference | 55.9 | +2.2 | #139 of 312, top 45% |
| Multilingual | 49.1 | +1.7 | #135 of 297, top 46% |
| Long Context | 41.6 | +0.7 | #135 of 296, top 46% |
Weakest categories
| Category | Score | vs median | Rank |
|---|---|---|---|
| Coding | 33.3 | −5.4 | #259 of 340, top 77% |
| Knowledge | 35.0 | −2.4 | #191 of 314, top 61% |
| Instruction Following | 69.6 | −1.7 | #171 of 305, top 57% |
Closest competitors
The models ranked just above and below GLM-4.5-Air. When scores are this close, price and speed are often the better way to choose.
| Model | Rank | Score | Blended $/M | Speed | |
|---|---|---|---|---|---|
| Gemini 2.0 Pro | #173 | 39.1 | — | — | Compare |
| Molmo 2 8b | #174 | 39.1 | — | — | Compare |
| Mercury 2 | #175 | 39.1 | $0.38 | — | Compare |
| Mistral Large 3 | #176 | 39.1 | $0.38 | 7 | Compare |
| MiniMax-M2.1 | #178 | 38.9 | $0.52 | — | Compare |
| Qwen3-30B-A3B | #179 | 38.9 | $0.21 | 42 | Compare |
| GLM-4.7-Flash | #180 | 38.8 | $0.15 | — | Compare |
| Qwen2.5 Plus 1127 | #181 | 38.8 | — | — | Compare |
Sponsored placements are available on pages like this one. Advertise on Noometry
Benchmark results
Every published result we track, with its source. Bold rows are the ones used for ranking; where several exist we prefer independent runs over self-reported numbers.
Coding
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| GSO | 2.9% | #29 of 31, top 94% | Epoch AI | ||
| LMArena Coding | 1397 | #138 of 294, top 47% | LMArena | 2026-10-08 |
Reasoning
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| Kagi LLM Benchmark | 43% | #73 of 99, top 74% | Kagi LLM Benchmark | ||
| LMArena Hard Prompts | 1379 | #139 of 297, top 47% | LMArena | 2026-10-08 | |
| ForecastBench | 59.2 | #40 of 72, top 56% | Epoch AI |
Math
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| Omni-MATH | 39.1% | #27 of 57, top 48% | HELM Capabilities | ||
| LMArena Math | 1396 | #129 of 285, top 46% | LMArena | 2026-10-08 |
Knowledge
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| Humanity's Last Exam | 8.1% | #27 of 41, top 66% | Epoch AI | ||
| MMLU-Pro | 76.2% | #26 of 58, top 45% | HELM Capabilities | ||
| Vectara Hallucination Rate (lower is better) | 9.3% | #44 of 96, top 46% | Vectara Hallucination Leaderboard | ||
| GPQA (HELM) | 59.4% | #24 of 57, top 43% | HELM Capabilities | ||
| LMArena Expert | 1370 | #142 of 273, top 53% | LMArena | 2026-10-08 |
Multilingual
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LMArena Non-English | 1366 | #135 of 297, top 46% | LMArena | 2026-10-08 | |
| LMArena Chinese | 1426 | #122 of 285, top 43% | LMArena | 2026-10-08 | |
| LMArena French | 1399 | #118 of 223, top 53% | LMArena | 2026-10-08 | |
| LMArena German | 1377 | #112 of 231, top 49% | LMArena | 2026-10-08 | |
| LMArena Japanese | 1348 | #99 of 211, top 47% | LMArena | 2026-10-08 | |
| LMArena Korean | 1308 | #126 of 213, top 60% | LMArena | 2026-10-08 | |
| LMArena Russian | 1373 | #134 of 283, top 48% | LMArena | 2026-10-08 | |
| LMArena Spanish | 1386 | #123 of 226, top 55% | LMArena | 2026-10-08 |
Instruction Following
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| IFEval | 81.2% | #37 of 57, top 65% | HELM Capabilities | ||
| LMArena Instruction Following | 1354 | #142 of 298, top 48% | LMArena | 2026-10-08 |
Long Context
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LMArena Longer Query | 1366 | #142 of 291, top 49% | LMArena | 2026-10-08 |
Writing & Preference
| Benchmark | Score | Position | Setting | Source | Date |
|---|---|---|---|---|---|
| LMArena Text | 1384 | #133 of 297, top 45% | LMArena | 2026-10-08 | |
| LMArena Creative Writing | 1343 | #135 of 295, top 46% | LMArena | 2026-10-08 | |
| WildBench | 78.9% | #37 of 57, top 65% | HELM Capabilities | ||
| LMArena Multi-Turn | 1371 | #138 of 295, top 47% | LMArena | 2026-10-08 |
API pricing by provider
| Route | Input $/M | Output $/M | Cached input $/M | Checked |
|---|---|---|---|---|
| openrouter | $0.13 | $0.85 | $0.025 | 2026-10-10 |
| zai | $0.20 | $1.10 | $0.03 | 2026-10-10 |
Compare GLM-4.5-Air
- GLM-4.5-Air vs Mistral Large 3
- GLM-4.5-Air vs MiniMax-M2.1
- GLM-4.5-Air vs Mercury 2
- GLM-4.5-Air vs Qwen3-30B-A3B
- GLM-4.5-Air vs Molmo 2 8b
- GLM-4.5-Air vs GLM-4.7-Flash
- GLM-4.5-Air vs GPT-6 Astra
- GLM-4.5-Air vs Claude Fable 5.1
- GLM-4.5-Air vs Gemini 3.8 Flash
- GLM-4.5-Air vs Kimi K3
- GLM-4.5-Air vs Grok 4.6
- GLM-4.5-Air vs Qwen3.8 Max
- GLM-4.5-Air vs Muse Spark 1.3
- GLM-4.5-Air vs DeepSeek V4 Pro
Other Z.ai (Zhipu) models
- GLM-5.354.8
- GLM-5.3-Flash51.8
- GLM-5.251.1
- GLM-5.147.8
- GLM-546.1
- GLM-5V-Turbo43.8
- GLM-4.542.0
- GLM-4.742.0
Frequently asked questions
How good is GLM-4.5-Air?
GLM-4.5-Air by Z.ai (Zhipu) ranks 177th of 354 ranked models on the Noometry Index as of October 2026, with a score of 38.9. Its strongest category is long context, where it ranks 135th. API pricing starts at $0.20 per million input tokens and $1.10 per million output tokens, with a 131K-token context window.
How much does GLM-4.5-Air cost?
GLM-4.5-Air costs $0.20 per million input tokens and $1.10 per million output tokens on Z.ai (Zhipu)'s own API, with cached input at $0.03.
What is GLM-4.5-Air's context window?
GLM-4.5-Air accepts up to 131K tokens of input and can write up to 98K tokens in one response.
Is GLM-4.5-Air open source?
Yes. GLM-4.5-Air's weights are downloadable from Hugging Face (zai-org/GLM-4.5-Air); check the license for commercial terms.
How fast is GLM-4.5-Air?
GLM-4.5-Air generated about 160 output tokens per second in the Kagi LLM Benchmark's timed runs. Speed varies by provider, load and reasoning effort.
What are GLM-4.5-Air's strengths and weaknesses?
Relative to other ranked models, GLM-4.5-Air places best in writing & preference, multilingual, long context and lowest in coding, knowledge, instruction following.
What is GLM-4.5-Air best at?
Its best category is long context, where it ranks 135th on Noometry.