[whatmodel.sh](../../README.md) / GLM-5 # GLM-5 Z.ai · released 2026-02 · 744B params (40B active) · mit **score 68.6** ±0.6 · #91 of 732 · from 59 benchmark results across 17 sources Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only. ## What it's good at ● marks GLM-5 among every model on that task or test, worst to best. 
task            vs the others         rank  counts  tests
agent tasks     ▁▁▃▆█▄▅●▄▄▂▁    #79 of 313     37%     18
coding          ▁▁▁▃▅▇█▇▅●▃▁    #75 of 469     21%      6
chat & writing  ▁▁▁▃▆▄▆█▆▇●▃   #46 of 402*    3.3%      1
math            ▁▁▃▅█▆▄●▄▂▁▁    #74 of 431     22%     13
web search      ▁▁▁▅●█▃▄·▁·▁     #58 of 84    5.4%      4
agent tasks · #79 of 313 · 37% of its score Beats [GPT-5](gpt-5.md), [Gemini 3 Flash (Preview)](gemini-3-flash-preview.md), [Grok 4.20 (0309 reasoning)](grok-4-20.md) here, though they rank higher overall. Trails [MiMo-V2.5 (310B-A15B)](mimo-v2-5.md), [MiMo-V2.5-Pro (1T-A42B)](mimo-v2-5-pro.md) here, though they rank lower overall. 
test                                result  vs the others in its run  difficulty            counts  why
SWE-bench Verified                   72.1%  ▁··▁·▁▂▁●█▄▁ beats 10/29  hard · avg 19%          9.0%  sharp · at its level · several runs
MCP-Atlas (public set) (maker)       67.8%  ▄······▄▄▄▄● beats 5/6    hard · avg 26%          5.1%  at its level · several runs
Terminal-Bench 2.0 (maker)           56.2%  ▂▂▅▅███●▆▂▃▃ beats 23/36  hard · avg 15%          2.8%  at its level · small run · several runs
SWE-bench Pro (split unspecified)…   55.1%  ▂·▁▁▁▁▂▃●▇▂▁ beats 34/58  very hard · avg 9.6%    2.1%  sharp · at its level · small run
BrowseComp (maker)                   62.0%  ▂·▃▁▂▂▆▅●▂▂▄ beats 29/46  hard · avg 25%          1.8%  at its level · small run · several runs
τ³-bench banking_knowledge (v1.0.…    9.8%  ●▃▂▃·▆▃▂█▅▅▂ beats 1/25   very hard · avg 4.3%    1.8%  typical
Toolathlon (Tool Decathlon) (make…   38.0%  ▂▂▄▂▄▄●▂█▆██ beats 10/27  very hard · avg 13%     1.5%  at its level · small run · several runs
BrowseComp with context managemen…   75.9%  ▄·▄▂·●▂·▆▄·▂ beats 5/15   medium · avg 46%        1.4%  small run · several runs
τ³-bench airline (tau2-bench ≥ v1…   82.5%  ▃······▃·▃●▃ beats 3/6    medium · avg 51%        1.4%  typical
SWE Atlas (Codebase Q&A, Test Wri…   24.0%  ▄·▄··●···▄·█ beats 3/6    very hard · avg 6.6%    1.3%  typical
BrowseComp-ZH (maker)                72.7%  ▁▁▁▁▁▃·▁▄▅▄● beats 23/23  hard · avg 25%          1.3%  at its level · small run
τ³-bench retail (tau2-bench ≥ v1.…   73.7%  ▄●·█··▄·▄··▄ beats 1/6    medium · avg 48%        1.2%  typical
Vibe Code Bench (test split, 50 w…   23.4%  ▃··▃·▃█●▃▃▃█ beats 8/14   very hard · avg 2.8%    1.2%  sharp · at its level · noisy
τ²-bench aggregate (maker)           89.7%  ▁▁▁·▁▁▃▃▃▄█● beats 38/39  medium · avg 54%        1.2%  small run
SWE-bench Multilingual (maker)       73.3%  ▁··▁▃▁▅█●▂▂▁ beats 43/65  hard · avg 28%          1.2%  at its level · small run
τ³-bench telecom (tau2-bench ≥ v1…   86.8%  ●·▄█▄······▄ beats 1/6    medium · avg 64%        1.1%  typical
Humanity's Last Exam with tools (…   50.4%  ▁·▁▂▂▁▄▄▇●▄▅ beats 36/62  hard · avg 16%          0.9%  at its level · small run
τ³-bench (maker)                     69.2%  ▁▁▂▁·▁▁·▁▁▂● beats 18/25  medium · avg 38%        0.6%  at its level · small run
coding · #75 of 469 · 21% of its score Beats [DeepSeek V4 Flash Vision (Exp)](deepseek-v4-flash-vision-exp.md), [GPT-5.2](gpt-5-2.md), [GPT-5](gpt-5.md) here, though they rank higher overall. Trails [Llama 3.1 Nemotron Ultra 253B v1](llama-3-1-nemotron-ultra-253b.md), [MiMo-V2.5 (310B-A15B)](mimo-v2-5.md), [MiMo-V2.5-Pro (1T-A42B)](mimo-v2-5-pro.md) here, though they rank lower overall. 
test                                result  vs the others in its run   difficulty            counts  why
SWE-bench Verified                   72.1%  ▁··▁·▁▂▁●█▄▁ beats 10/29   hard · avg 19%          9.0%  sharp · at its level · several runs
Arena WebDev (Code Arena)           1435.6  ▁▁▃▂▅●▃▆▅▅▁▂ beats 56/129  avg 981                 5.9%  sharp
SWE-bench Pro (split unspecified)…   55.1%  ▂·▁▁▁▁▂▃●▇▂▁ beats 34/58   very hard · avg 9.6%    2.1%  sharp · at its level · small run
SWE Atlas (Codebase Q&A, Test Wri…   24.0%  ▄·▄··●···▄·█ beats 3/6     very hard · avg 6.6%    1.3%  typical
Vibe Code Bench (test split, 50 w…   23.4%  ▃··▃·▃█●▃▃▃█ beats 8/14    very hard · avg 2.8%    1.2%  sharp · at its level · noisy
SWE-bench Multilingual (maker)       73.3%  ▁··▁▃▁▅█●▂▂▁ beats 43/65   hard · avg 28%          1.2%  at its level · small run
chat & writing · #46 of 402 · 3.3% of its score Beats [MiMo-V2.6-Flash (309B-A15B)](mimo-v2-6-flash.md), [Claude Haiku 5.5](claude-haiku-5-5.md), [GPT-6 Luna](gpt-6-luna.md) here, though they rank higher overall. Trails [MiMo-V2.5-Pro (1T-A42B)](mimo-v2-5-pro.md) here, though they rank lower overall. 
test            result  vs the others in its run    difficulty  counts  why
LMArena (text)  1457.2  ▁▁▁▂▃▃▃▆▆▇●▃ beats 332/385  avg 1260      3.3%  blunt
math · #74 of 431 · 22% of its score Beats [Claude Sonnet 4.6](claude-sonnet-4-6.md), [Claude Opus 4.5](claude-opus-4-5.md), [Qwen3.7 Plus](qwen3-7-plus.md) here, though they rank higher overall. Trails [Granite 4.2 30B](granite-4-2-30b.md), [INTELLECT-3](intellect-3.md), [Qwen3 235B-A22B Thinking 2507](qwen3-235b-a22b-thinking-2507.md) here, though they rank lower overall. 
test                                result  vs the others in its run    difficulty            counts  why
MathArena ArXivMath February 2026…   41.4%  ▁▁▁▁█●▄▃▁▃·▃ beats 14/24    very hard · avg 9.4%    3.8%  at its level
MathArena ArXivMath March 2026 (3…   45.8%  ●▃▅·█▃▃·▅·▃▃ beats 2/14     hard · avg 16%          3.2%  at its level
MathArena ArXivMath January 2026 …   54.3%  ▂··▄▂▄●▆▆▄▄█ beats 9/23     hard · avg 23%          2.8%  at its level
AIME 2026 (maker)                    95.4%  ▁▁▁▂▂▂▆█●▂·▁ beats 43/52    medium · avg 63%        2.0%  sharp · maxed out · small run
HMMT February 2026 (maker)           82.8%  ▃▃▂▅·███●▅▃▅ beats 24/35    hard · avg 30%          2.0%  sharp · small run
MathArena ArXivMath December 2025…   38.2%  ▃·▃●█▁·▄·▁▁▁ beats 5/18     hard · avg 19%          1.9%  at its level
IMO-AnswerBench (maker)              82.5%  ▁▁·▁▁▁▄▆●▄▂▃ beats 30/44    medium · avg 38%        1.5%  small run
OTIS mock AIME 2024–2025             80.0%  ▃▃▄▃▃▃▇●█▃▂▅ beats 113/197  very hard · avg 7.2%    1.4%  sharp · noisy
HMMT February 2025 (maker)           97.9%  ▁▁▁▃▁▅▅▃██▁● beats 29/31    easy · avg 80%          1.3%  maxed out · small run
MathArena BrokenArXiv March 2026 …   12.1%  █·●█▂·▂▂···▂ beats 6/14     very hard · avg 1.0%    0.9%  sharp · noisy
MathArena BrokenArXiv February 20…   11.7%  ▃▂▂▃●▃▂·▃··▂ beats 8/16     very hard · avg 0.9%    0.8%  sharp · noisy
HMMT November 2025 (maker)           96.9%  ▁▁····▁▅▃█▇● beats 23/23    easy · avg 77%          0.8%  maxed out · small run
AIME 2026 I (maker)                  92.7%  ●··········█ beats 0/1                                 —  not counted yet: no other model in the …
web search · #58 of 84 · 5.4% of its score Beats [Kimi K2.6](kimi-k2-6.md), [GPT-5 Pro](gpt-5-pro.md), [DeepSeek V4 Flash (Preview, Apr 2026)](deepseek-v4-flash-preview.md) here, though they rank higher overall. Trails [GPT-4o mini (2024-07-18)](gpt-4o-mini-2024-07-18.md), [Mistral Small 3.1 24B (2503)](mistral-small-3-1.md), [Llama 4 Scout (109B-A17B)](llama-4-scout.md) here, though they rank lower overall. 
test                                result  vs the others in its run  difficulty        counts  why
BrowseComp (maker)                   62.0%  ▂·▃▁▂▂▆▅●▂▂▄ beats 29/46  hard · avg 25%      1.8%  at its level · small run · several runs
BrowseComp with context managemen…   75.9%  ▄·▄▂·●▂·▆▄·▂ beats 5/15   medium · avg 46%    1.4%  small run · several runs
BrowseComp-ZH (maker)                72.7%  ▁▁▁▁▁▃·▁▄▅▄● beats 23/23  hard · avg 25%      1.3%  at its level · small run
Humanity's Last Exam with tools (…   50.4%  ▁·▁▂▂▁▄▄▇●▄▅ beats 36/62  hard · avg 16%      0.9%  at its level · small run
other tests (16) 
test                                result  vs the others in its run    difficulty            counts  why
GPQA Diamond                         87.8%  ▁▁▁▁▂▄▆▄▅▇●▅ beats 159/211  hard · avg 39%          9.2%  sharp · several runs
Humanity's Last Exam (maker)         30.5%  ▁▁▁▄▆▆▇█●▆▁▃ beats 77/111   very hard · avg 5.8%    2.6%  at its level · small run · several runs
KINA (Knowledge Index of Noah's A…   35.9%  ▁▁▂▁▃▄█▆●▅▂▃ beats 28/39    very hard · avg 21%     2.5%  at its level
ARC-AGI-2 public evaluation set (…    5.4%  █··▂●▁▃▁▃·▁▃ beats 12/24    very hard · avg 0.3%    2.3%  sharp · noisy
ARC-AGI-1 public evaluation set (…   58.6%  ▁·▇▃▁·●▇▁·▇▃ beats 11/27    very hard · avg 4.0%    2.3%  sharp · at its level · noisy
EQ-Bench 3 (emotional intelligenc…  1307.3  ▁·▂▁▄▅▅▃●▄▁▁ beats 49/62    avg 1034                2.0%  typical
Chess Puzzles (Epoch)                10.0%  █▁▁▁●▂▃▃▃▂▁▁ beats 55/137   very hard · avg 5.8%    1.8%  sharp · noisy
WildClawBench                        42.6%  ▃·▅▅█·▅●▅▃▃▃ beats 10/17    very hard · avg 11%     1.6%  at its level
AraGen-v3 3C3H score (Inception, …   46.0%  ▁▁·▁▂▂▄█●▅▄▄ beats 40/66    hard · avg 29%          1.5%  at its level
LongBench v2 (maker)                 64.5%  ▂·▁·▂▃▁▄▄▃█● beats 31/33    very hard · avg 35%     1.3%  at its level · small run
LongSocialBench (full-context, 4-…   41.5%  ▆▂·▄●·▂▂·▄▄▂ beats 7/16     very hard · avg 36%     1.3%  blunt · at its level
CyberGym (maker)                     48.3%  ▄···●··▄█▄█▄ beats 1/8      hard · avg 19%          1.1%  at its level · small run · several runs
NL2Repo (maker)                      35.9%  ▂▂▂·▃▃●█▃▂▅▃ beats 8/22     very hard · avg 11%     1.0%  at its level · small run
SOOHAK Challenge (340 problems, A…    9.6%  █·█●███··█·█ beats 2/7      very hard · avg 4.3%    0.5%  typical
SOOHAK-Mini (702 problems, Avg@3)    63.1%  ▄··▄·▄·▄●█·▄ beats 4/7      medium · avg 36%        0.4%  blunt · at its level
Vectara hallucination leaderboard…   89.9%  ▁▂▁▂▅●▂▃▄▃▁▁ beats 44/98    very easy · avg 90%     0.0%  blunt
## 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
zai-org/GLM-5.1 model card (submitted /…       13   17% ▇▇
Epoch AI                                        4   14% ▇▇
MathArena ArXivMath / BrokenArXiv / ArX…        6   13% ▇▇
GLM-5 Team (submitted / maker-reported)        10   13% ▇▇
zai-org/GLM-5 model card (submitted / m…        8   10% ▇
LMArena leaderboard dataset                     2    9% ▇
τ²/τ³-bench leaderboard: Sierra's own e…        4    6% ▇
ARC Prize Foundation model_baseline out…        2    5% ▇
Knowledge Index of Noah's Ark                   1    3% ▇
EQ-Bench 3 canonical leaderboard results        1    2% ▇
WildClawBench: A Benchmark for Real-Wor…        1    2% ▇
Inception Arabic/Hindi leaderboards             1    2% ▇
SWE Atlas: Benchmarking Coding Agents B…        1    1% ▇
LongSocialBench: Do Long-Context LLMs U…        1    1% ▇
Vibe Code Bench                                 1    1% ▇
Soohak: A Mathematician-Curated Benchma…        2    1% ▇
Vectara hallucination leaderboard               1    0% ▇
prior: pull toward the middle                        0% ▇
weighted average                                   = 68.6
Licences: zai-org/GLM-5.1 model card ([mit](../../licenses/zai-org--GLM-5.1.txt)) · Epoch AI ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · MathArena ArXivMath / BrokenArXiv / ArXivLean model-output datasets ([CC BY-SA 4.0](../../licenses/matharena-sharealike-NOTICE.txt)) · GLM-5 Team ([CC BY 4.0](../../licenses/report-2602.15763v2.txt)) · zai-org/GLM-5 model card ([MIT](../../licenses/zai-org--GLM-5.txt)) · LMArena leaderboard dataset ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · τ²/τ³-bench leaderboard: Sierra's own evaluation runs ([MIT](https://opensource.org/license/mit)) · ARC Prize Foundation model\_baseline outputs on ARC-AGI-1/2 public evaluation sets ([MIT](../../licenses/arcprize-model-baseline-outputs-MIT.txt)) · Knowledge Index of Noah's Ark ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · EQ-Bench 3 canonical leaderboard results ([MIT](../../licenses/eqbench3-MIT.txt)) · WildClawBench: A Benchmark for Real-World ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · Inception Arabic/Hindi leaderboards ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt)) · SWE Atlas: Benchmarking Coding Agents Beyond Issue Resolution ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · LongSocialBench: Do Long-Context LLMs Understand Online Discussion Threads? ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)) · Vibe Code Bench ([CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/)) · Soohak: A Mathematician-Curated Benchmark for Evaluating Research-level Math Capabilities of LLMs ([CC BY-SA 4.0](../../licenses/paper-2605.09063v3-CC-BY-SA-4.0.txt)) · Vectara hallucination leaderboard ([Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0))
Not counted (1) 
source                  test         score  why
zai-org/GLM-5 model c…  AIME 2026 I  92.7%  too few models have it yet
## How to run it ### Through an API api (Z.AI) · $1.55/M blended · $1 in / $3.2 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-5.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-5.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-5 (ud-q4_k_xl, needs ~446.9 GB, from [huggingface.co/unsloth](https://huggingface.co/unsloth/GLM-5-GGUF/tree/main)) ```sh llama cli -hf unsloth/GLM-5-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-5-GGUF:UD-Q4_K_XL ``` Other versions: ud-q2_k_xl (~223.7 GB) `llama cli -hf unsloth/GLM-5-GGUF:UD-Q2_K_XL` · ud-iq1_s (~112.1 GB) `llama cli -hf unsloth/GLM-5-GGUF:UD-IQ1_S` ### 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-5 (ud-q4_k_xl, needs ~446.9 GB, from [huggingface.co/unsloth](https://huggingface.co/unsloth/GLM-5-GGUF/tree/main)) ```sh lms get https://huggingface.co/unsloth/GLM-5-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 (~223.7 GB) `lms get https://huggingface.co/unsloth/GLM-5-GGUF@UD-Q2_K_XL -y` · ud-iq1_s (~112.1 GB) `lms get https://huggingface.co/unsloth/GLM-5-GGUF@UD-IQ1_S -y` · 4bit (~418.6 GB, Mac only) `lms get https://huggingface.co/mlx-community/GLM-5-4bit -y` ### 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-5 (4bit, needs ~418.6 GB, Mac only, from [huggingface.co/mlx-community](https://huggingface.co/mlx-community/GLM-5-4bit/tree/97b2af2010d640c7c0fa441f0e665e5092c16c07)) ```sh mlx_lm.chat --model mlx-community/GLM-5-4bit ``` 3. Or start an OpenAI-compatible server at http://localhost:8080/v1 ```sh mlx_lm.server --model mlx-community/GLM-5-4bit ``` Model repository and documentation: [huggingface.co/zai-org](https://huggingface.co/zai-org/GLM-5) [interactive version](https://whatmodel.sh/model/glm-5) · [JSON](https://whatmodel.sh/api?format=json&all=1)