[whatmodel.sh](../../README.md) / GLM-4.5-Air # GLM-4.5-Air Z.ai · released 2025-07 · 106B params (12B active) · mit **score 62.2** ±0.7 · #177 of 732 · from 34 benchmark results across 10 sources Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only. ## What it's good at ● marks GLM-4.5-Air among every model on that task or test, worst to best. 
task            vs the others          rank  counts  tests
agent tasks     ▁▁▃▆█●▅▅▄▄▂▁    #152 of 313     20%      9
coding          ▁▁▁▃▅▇█●▅▄▃▁    #165 of 469     11%      4
chat & writing  ▁▁▁▃▆▄▆█●▇▇▃   #140 of 402*    6.9%      2
math            ▁▁▃▅█▆▄●▄▂▁▁     #99 of 431     18%      3
web search      ●▁▁▅▄█▃▄·▁·▁     #84 of 84*    1.6%      1
agent tasks · #152 of 313 · 20% of its score Beats [o1 (2024-12-17)](o1-2024-12-17.md), [o4-mini](o4-mini.md), [Qwen3 235B-A22B Thinking 2507](qwen3-235b-a22b-thinking-2507.md) here, though they rank higher overall. Trails [SmolLM3 3B](smollm3-3b.md), [Jamba 1.5 Large](jamba-1-5-large.md), [GPT-4o (2024-05-13)](gpt-4o-2024-05-13.md) here, though they rank lower overall. 
test                                result  vs the others in its run    difficulty           counts  why
SWE-bench Verified                   57.4%  ▁▁▁▁▁▁▂●█▆▁▁ beats 51/138   hard · avg 23%         7.4%  sharp · at its level · small run · seve…
τ²-bench Telecom                     28.1%  ▁·▅●▆▅▂▃▃▂▃▁ beats 22/89    very hard · avg 15%    2.6%  sharp · at its level · noisy · small run
τ-bench Retail (maker)               77.9%  ▁····▃·▂▇█▂● beats 24/28    medium · avg 46%       2.4%  at its level · small run
τ-bench Airline (maker)              60.8%  ▁▁▁·▅▄▁█▇▁▁● beats 28/28    hard · avg 30%         1.7%  at its level · small run
BrowseComp (maker)                   21.3%  ▂·▃▁●▂▆▅█▂▂▄ beats 7/46     very hard · avg 13%    1.6%  at its level · small run
BFCL V4 (2025-11-03 archive, mode…   68.9%  ▃··▅▃▃·▅·▃·● beats 8/10     hard · avg 33%         1.3%  at its level
Terminal-Bench (pre-2.0 release) …   30.0%  ▃·▃···▃▃●·▅█ beats 4/10     hard · avg 15%         1.3%  at its level · small run
Galileo Agent Leaderboard v2 (act…   44.0%  ▅▃·▃▃▅▅●▃█▅▅ beats 11/19    hard · avg 26%         1.2%  at its level
BFCL v3 (maker)                      76.4%  ▁▁▁▁▁▃▅█▆▅▄● beats 168/171  medium · avg 60%       0.7%  blunt · at its level · small run
coding · #165 of 469 · 11% of its score Beats [Qwen3-Coder-Next (80B-A3B)](qwen3-coder-next.md), [Gemini 2.5 Pro](gemini-2-5-pro.md), [Gemini 3.1 Flash-Lite](gemini-3-1-flash-lite.md) here, though they rank higher overall. Trails [QwQ-32B-Preview](qwq-32b-preview.md), [Ministral 3 8B Reasoning 2512](ministral-3-8b-reasoning.md), [Phi-4-reasoning 14B](phi-4-reasoning.md) here, though they rank lower overall. 
test                                result  vs the others in its run   difficulty        counts  why
SWE-bench Verified                   57.4%  ▁▁▁▁▁▁▂●█▆▁▁ beats 51/138  hard · avg 23%      7.4%  sharp · at its level · small run · seve…
LiveCodeBench v6 (Aug 2024–May 20…   61.5%  ●▄·▄·····█·▄ beats 0/5     medium · avg 45%    2.0%  at its level
SciCode (level unspecified) (make…   37.3%  ▁····▃▃▄▅●▆▂ beats 23/40   hard · avg 18%      1.4%  at its level · small run
LiveCodeBench 2024-07 to 2025-01 …   70.7%  ●··········█ beats 0/1                            —  not counted yet: no other model in the …
chat & writing · #140 of 402 · 6.9% of its score Beats [o3-mini](o3-mini.md), [INTELLECT-3](intellect-3.md), [MiniMax-M2](minimax-m2.md) here, though they rank higher overall. Trails [LFM2.5 8B-A1B](lfm2-5-8b-a1b.md), [LFM2.5 2.6B](lfm2-5-2-6b.md), [Nemotron-Cascade-8B](nemotron-cascade-8b.md) here, though they rank lower overall. 
test                                result  vs the others in its run    difficulty      counts  why
LMArena (text)                      1373.0  ▁▁▁▂▃▃▃▆●▇█▃ beats 214/385  avg 1260          5.8%  blunt
IFEval prompt-level strict (maker)   86.3%  ▁▁··▂··▂▇●▃▃ beats 20/26    easy · avg 74%    1.1%  small run
math · #99 of 431 · 18% of its score Beats [Claude Sonnet 4.6](claude-sonnet-4-6.md), [MiniMax-M3](minimax-m3.md), [Nemotron 3 Ultra 550B-A55B](nemotron-3-ultra-550b-a55b.md) here, though they rank higher overall. Trails [Gemini 2.5 Flash](gemini-2-5-flash.md) here, though they rank lower overall. 
test       result  vs the others in its run  difficulty           counts  why
AIME 2024   84.6%  █···█●█·█··█ beats 2/5    medium · avg 50%        10%  sharp · several runs
AIME 2025   82.0%  █·█···●██··█ beats 2/5    medium · avg 42%       5.6%  sharp
MATH-500    97.8%  ▄······▄▄▄·● beats 4/5    very easy · avg 92%    2.6%  maxed out · small run · several runs
web search · #84 of 84 · 1.6% of its score Trails [GPT-4.1 nano](gpt-4-1-nano-2025-04-14.md), [GPT-4o mini (2024-07-18)](gpt-4o-mini-2024-07-18.md), [Mistral Small 3.1 24B (2503)](mistral-small-3-1.md) here, though they rank lower overall. 
test                result  vs the others in its run  difficulty           counts  why
BrowseComp (maker)   21.3%  ▂·▃▁●▂▆▅█▂▂▄ beats 7/46   very hard · avg 13%    1.6%  at its level · small run
other tests (14) 
test                                result  vs the others in its run    difficulty            counts  why
GPQA Diamond                         73.3%  █··●█·██···█ beats 1/5      hard · avg 41%           12%  sharp · at its level
Medmarks-V mean win rate (19 veri…  0.5410  ▁▄▃▃▇▄█●▅▁▁▁ beats 23/30    avg 0.418                11%  sharp
Humanity's Last Exam                 13.3%  ▄····●·█·▄·▄ beats 1/5      very hard · avg 4.1%    7.2%  several runs
MMLU-Pro                             73.9%  ▁▁·▁▁▁▃▂▄▅●▁ beats 134/188  medium · avg 50%        5.5%  at its level · several runs
Medmarks-V weighted mean win rate…   54.1%  ▂▃▅▂▆▃█▅●▅▃▆ beats 24/34    medium · avg 41%        4.7%  blunt · at its level
GPQA (subset unspecified) (maker)    75.0%  ▂▁▂▂▄▃▆▃▅●▃▃ beats 88/106   hard · avg 46%          3.6%  sharp · at its level · small run
MMLU (maker)                         87.4%  ▁▁▁▁▁▂▂▅██▆● beats 200/214  medium · avg 69%        2.8%  small run
SimpleQA (maker)                     14.5%  ▄▄▄▆█▆●██·▂█ beats 16/32    very hard · avg 7.5%    1.7%  small run
MultiChallenge (maker)               42.5%  ▁·▁·▁▃●▅▃▂█▅ beats 7/30     hard · avg 27%          1.2%  at its level · small run
EuroEval HellaSwag (English), tes…   80.4%  ▃·▁▁▂▁▂▅█●▄▁ beats 73/100   medium · avg 67%        0.6%  sharp · noisy
EuroEval Life in the UK (English …   95.8%  ▁▁▁▁▁▁▁▂▄█▄● beats 144/146  easy · avg 82%          0.3%  noisy
Vectara hallucination leaderboard…   98.1%  ▁▁▂▃▄▄█▆▃●▂▁ beats 123/140  very easy · avg 94%     0.2%  blunt · maxed out · noisy
Vectara hallucination leaderboard…   90.7%  ▁▂▁▂▅●▂▃▄▃▁▁ beats 54/98    very easy · avg 90%     0.0%  blunt
SysBench (ISR) (maker)               77.4%  ●··········█ beats 0/1                                 —  not counted yet: no other model in the …
## How the score was worked out Each result implies a score on its own, after adjusting for how hard that test is and who ran it. The model's score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its *counts* column above; here it is added up by who ran the tests. [Method](../method.md) 
source                                    results  share of the score
Prime Intellect INTELLECT-3 model card:…        7   35% ▇▇▇▇
GLM-4.5 Team (submitted / maker-reporte…       16   33% ▇▇▇▇
Medmarks: A Comprehensive Open-Source L…        2   15% ▇▇
LMArena leaderboard dataset                     1    6% ▇
SERA: Soft-Verified Efficient Repositor…        1    4% ▇
Solar Open Technical Report                     1    3% ▇
Berkeley Function Calling Leaderboard r…        1    1% ▇
Galileo Agent Leaderboard                       1    1% ▇
EuroEval leaderboards                           2    1% ▇
Vectara hallucination leaderboard               2    0% ▇
prior: pull toward the middle                        0% ▇
weighted average                                   = 62.2
Licences: Prime Intellect INTELLECT-3 model card: Evaluation table ([MIT](https://huggingface.co/PrimeIntellect/INTELLECT-3/blob/ff39d4a4688989f3f28868923d030c28e1b7d81c/README.md)) · GLM-4.5 Team ([CC BY 4.0](../../licenses/report-2508.06471v1.txt)) · Medmarks: A Comprehensive Open-Source LLM Benchmark Suite for Medical Tasks ([CC BY 4.0](../../licenses/paper-2605.01417v1-CC-BY-4.0.txt)) · LMArena leaderboard dataset ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · SERA: Soft-Verified Efficient Repository Agents ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · Solar Open Technical Report ([CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/)) · Berkeley Function Calling Leaderboard result archive ([Apache-2.0](../../licenses/bfcl-result-archive-v4-Apache-2.0.txt)) · Galileo Agent Leaderboard ([Apache-2.0](../../licenses/galileo-agent-leaderboard-Apache-2.0.txt)) · EuroEval leaderboards ([MIT](https://github.com/EuroEval/leaderboards/blob/1a77f1d2f8ba1ba05ab73287fb090d76bf41b2e0/LICENSE)) · Vectara hallucination leaderboard ([Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0))
Not counted (2) 
source        test                            score  why
GLM-4.5 Team  LiveCodeBench 2024-07 to 2025…  70.7%  too few models have it yet
GLM-4.5 Team  SysBench (ISR)                  77.4%  too few models have it yet
## How to run it ### Through an API api (Z.AI) · $0.43/M blended · $0.2 in / $1.1 out per 1M tokens · [docs.z.ai](https://docs.z.ai/guides/overview/pricing) [price source](https://github.com/anomalyco/models.dev/blob/a668908cfe7073faa0363699486fc41a3ff52412/providers/zai/models/glm-4.5-air.toml): Standard base token rate from models.dev. Provider-specific tiers, long-context surcharges, caching and promotions may differ. · [context source](https://github.com/anomalyco/models.dev/blob/a668908cfe7073faa0363699486fc41a3ff52412/models/zhipuai/glm-4.5-air.toml) ### llama.cpp (recommended) 1. Install llama.cpp. Homebrew and winget installs name the commands llama-cli and llama-server instead. [Other ways to install](https://github.com/ggml-org/llama.cpp/blob/master/docs/install.md) ```sh # macOS / Linux curl -LsSf https://llama.app/install.sh | sh # Windows irm https://llama.app/install.ps1 | iex ``` 2. Run GLM-4.5-Air (ud-q4_k_xl, needs ~64.1 GB, from [huggingface.co/unsloth](https://huggingface.co/unsloth/GLM-4.5-Air-GGUF/tree/main)) ```sh llama cli -hf unsloth/GLM-4.5-Air-GGUF:UD-Q4_K_XL ``` 3. Or start a local server and open the address it prints (http://localhost:9931) ```sh llama serve -hf unsloth/GLM-4.5-Air-GGUF:UD-Q4_K_XL ``` Other versions: ud-q2_k_xl (~32.3 GB) `llama cli -hf unsloth/GLM-4.5-Air-GGUF:UD-Q2_K_XL` ### LM Studio 1. Install LM Studio. Open the app once before using the lms command. [Download it (macOS / Windows / Linux)](https://lmstudio.ai/download) 2. Run GLM-4.5-Air (ud-q4_k_xl, needs ~64.1 GB, from [huggingface.co/unsloth](https://huggingface.co/unsloth/GLM-4.5-Air-GGUF/tree/main)) ```sh lms get https://huggingface.co/unsloth/GLM-4.5-Air-GGUF@UD-Q4_K_XL -y ``` 3. Then run lms ls to see its name, and lms chat followed by that name Other versions: ud-q2_k_xl (~32.3 GB) `lms get https://huggingface.co/unsloth/GLM-4.5-Air-GGUF@UD-Q2_K_XL -y` · 4bit (~60.1 GB, Mac only) `lms get https://huggingface.co/lmstudio-community/GLM-4.5-Air-MLX-4bit -y` · 8bit (~113.6 GB, Mac only) `lms get https://huggingface.co/lmstudio-community/GLM-4.5-Air-MLX-8bit -y` ### Ollama 1. Install Ollama. [Download it (macOS / Windows)](https://ollama.com/download) ```sh # Linux curl -fsSL https://ollama.com/install.sh | sh ``` 2. Run GLM-4.5-Air (ud-q2_k_xl, needs ~32.3 GB, from [huggingface.co/unsloth](https://huggingface.co/unsloth/GLM-4.5-Air-GGUF/tree/main)) ```sh ollama run hf.co/unsloth/GLM-4.5-Air-GGUF:UD-Q2_K_XL ``` ### MLX (Mac) 1. Install MLX (Mac). Apple silicon Macs only, text only. ```sh # with uv uv tool install mlx-lm # or pip install mlx-lm ``` 2. Run GLM-4.5-Air (4bit, needs ~60.1 GB, Mac only, from [huggingface.co/lmstudio-community](https://huggingface.co/lmstudio-community/GLM-4.5-Air-MLX-4bit/tree/92ce339f9fe1bbb3afa840d46cad1988b6a0e0bd)) ```sh mlx_lm.chat --model lmstudio-community/GLM-4.5-Air-MLX-4bit ``` 3. Or start an OpenAI-compatible server at http://localhost:8080/v1 ```sh mlx_lm.server --model lmstudio-community/GLM-4.5-Air-MLX-4bit ``` Other versions: 8bit (~113.6 GB, Mac only) `mlx_lm.chat --model lmstudio-community/GLM-4.5-Air-MLX-8bit` Model repository and documentation: [huggingface.co/zai-org](https://huggingface.co/zai-org/GLM-4.5-Air) ## Published speed measurements Reference results for the hardware, engine and settings below. These are not verified speeds of the downloads above or predictions for your machine. 
hardware                      precision         speed  engine and settings
AMD Ryzen AI Max+ 395 (Stri…  UD-Q4_K_XL  23.35 tok/s  llama.cpp 7ad67ba9f (commit from …
Each speed links to its source, which has the full settings. [interactive version](https://whatmodel.sh/model/glm-4-5-air) · [JSON](https://whatmodel.sh/api?format=json&all=1)