GLM-4.5

+ contribute

GLM-4.5 is an open-weights language model from Z.ai, released July 2025. Its published size is 355B parameters (32B active per token) and a 131k-token context window. Licence: mit.

64.7±0.6#147 of 719 LLMsfrom 52 benchmark results across 9 sources

$ how it compares

The LLMs ranked either side of GLM-4.5, out of 828 with a score, then more from Z.ai.

#	model	score	where	RAM	released
167	INTELLECT-3 Prime Intellect	65±1	local	~48GB	2025-11
168	Nemotron 3 Super 120B-A12B NVIDIA	65±1	api $0.28/M +local	~73GB	2026-03
169	GLM-4.5 Z.ai	65±1	api $1.00/M +local	~58GB	2025-07
170	Solar Pro 4 Upstage	65±2	api $0.53/M		2026-08
171	MiniMax-M2 MiniMax	65±1	api $0.53/M +local	~75GB	2025-10
	more from Z.ai
33	GLM-5.3 Z.ai	76±1	api $2.15/M +local	~239GB	2026-08
52	GLM-5.3-Flash Z.ai	74±1	api $0.24/M +local	~103GB	2026-08
57	GLM-5.2 Z.ai	74±1	api $2.15/M +local	~121GB	2026-06

see all LLMs · find the ones that fit your machine

$ what it's good at

agent tasks▁▁▃▆█▆▅▅▅▄▃▁ #126 of 300 · 18% of its score · 11 tests

Beats o3, Gemini 2.5 Pro, Gemini 3.1 Flash-Lite here, though they rank higher overall. Trails SmolLM3 3B, Jamba 1.5 Large, Nova 2 Pro (Preview) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
SWE-bench Verified lab +1	64.2% ✎	▁▁▁▁▁▁▂▄█▆▁▁ beats 61/131	hard · avg 24%	▇▇ 4.2%	sharp · at its level · small run · several runs
τ²-bench Retail	69.3% ✎	█···█·█·██·█ beats 2/5	medium · avg 36%	▇ 2.8%	at its level
BrowseComp lab +1	26.4% ✎	▂·▄▁▃▃▆▄█▂▃▄ beats 8/42	very hard · avg 12%	▇ 2.3%	at its level · small run · several runs
Terminal-Bench (pre-2.0 release) lab +1	37.5% ✎	▃·▃···▃▃▅·▅█ beats 7/10	very hard · avg 15%	▇ 1.9%	at its level · small run · several runs
τ²-bench Telecom	56.1% ✎	█··██·█·█··█ beats 3/5	very hard · avg 10%	▇ 1.7%	sharp · at its level · noisy
τ-bench Retail lab	79.7% ✎	▁····▃·▂▇█▂▆ beats 25/28	medium · avg 45%	▇ 1.6%	small run
τ²-bench Airline	66.0% ✎	▄·····█··▄·█ beats 4/5	hard · avg 29%	▇ 1.3%	at its level
τ-bench Airline lab	60.4% ✎	▁▁▁·▅▄▁█▇▁▁▇ beats 27/28	hard · avg 29%	▇ 1.1%	at its level · small run

3 more tests

test	result	vs the others in its run	difficulty	counts	why
BFCL V4 (2025-11-03 archive, models absent from the December archive)	72.0% ✎	▃··▅▃▃·▅·▃·█ beats 10/10	hard · avg 34%	▇ 0.8%	at its level
BFCL v3 +1	79.1% ✎	▁▁▁▁▁▃▅█▆▅▄▃ beats 170/170	medium · avg 60%	▇ 0.7%	blunt · at its level · small run · several runs
τ²-bench overall (weighted) lab	67.5% ✎	█··········█ beats 0/1		—	not counted yet: no other model in the same run to compare it with

coding▁▁▁▃▆██▇▆▄▃▁ #145 of 458 · 10% of its score · 6 tests

Beats Gemini 2.5 Pro, Gemini 3.1 Flash-Lite, Qwen3-Coder-Next (80B-A3B) here, though they rank higher overall. Trails Phi-4-reasoning 14B, DeepSeek V3 0324, Devstral Small 2 24B here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
SWE-bench Verified lab +1	64.2% ✎	▁▁▁▁▁▁▂▄█▆▁▁ beats 61/131	hard · avg 24%	▇▇ 4.2%	sharp · at its level · small run · several runs
LiveCodeBench (version unspecified)	61.1% ✎	▄······▄█▄·▄ beats 0/5	medium · avg 39%	▇ 2.3%	at its level
LiveCodeBench v6 (Aug 2024–May 2025)	64.5% ✎	▄▄·▄·····█·▄ beats 2/5	medium · avg 45%	▇ 1.3%	at its level
LiveCodeBench v6 (window unspecified) lab	63.3% ✎	▂▃▄▄▅██▅▄▇▃▆ beats 27/65	medium · avg 40%	▇ 1.1%	at its level · small run
SciCode (level unspecified) lab	41.7% ✎	▁····▃▄▄▄█▆▂ beats 26/34	hard · avg 19%	▇ 0.9%	at its level · small run
LiveCodeBench 2024-07 to 2025-01 (version not stated) lab	72.9% ✎	█··········█ beats 1/1		—	not counted yet: no other model in the same run to compare it with

chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #101 of 393 · 5.3% of its score · 2 tests

Beats LongCat 2.0 (1.6T-A48B), Inkling Small (266B-A12B), GPT-5.4 nano here, though they rank higher overall. Trails Mistral Large 3 (675B-A41B), GPT-4.1, Gemma 4 12B here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
LMArena (text)	1411.3 ✎	▁▁▁▂▃▃▃▆▆▇█▃ beats 257/379	avg 1268	▇ 3.4%	blunt
IFEval prompt-level strict +1	85.4% ✎	███··█·█···█ beats 0/5	easy · avg 74%	▇ 1.9%	several runs

math▁▂▃▅▆▅▄█▄▁▁▁ #111 of 398 · 27% of its score · 4 tests

Beats Claude Sonnet 4.6, MiniMax-M3, MiMo-V2.5 (310B-A15B) here, though they rank higher overall. Trails Gemini 2.5 Flash, Qwen3-VL 32B Thinking, gpt-oss-120b here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
AIME 2024 +2	85.8% ✎	█···███·█··█ beats 3/5	medium · avg 60%	▇▇▇▇ 10%	sharp · several runs
AIME 2025 +2	83.3% ✎	█·█···███··█ beats 3/5	medium · avg 51%	▇▇▇▇ 10%	sharp · several runs
HMMT 2025 (edition unspecified)	76.3% ✎	▄··▄··▄·▄··█ beats 1/5	hard · avg 29%	▇ 3.7%	sharp · at its level
MATH-500 +2	97.0% ✎	▄······▄▄▄·█ beats 3/5	very easy · avg 91%	▇ 2.9%	maxed out · small run · several runs

web search▃▁▂▆▄█▃▃▁▁·▁ #78 of 80 * · 2.3% of its score · 1 test

Trails GPT-4.1 nano, GPT-4o mini (2024-07-18), Mistral Small 3.1 24B (2503) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
BrowseComp lab +1	26.4% ✎	▂·▄▁▃▃▆▄█▂▃▄ beats 8/42	very hard · avg 12%	▇ 2.3%	at its level · small run · several runs

other tests17 not in a task · 44% of its score

test	result	vs the others in its run	difficulty	counts	why
GPQA Diamond +1	77.0% ✎	█··██·██···█ beats 3/5	hard · avg 42%	▇▇▇▇▇ 13%	sharp · at its level · several runs
Humanity's Last Exam +2	14.8% ✎	▄····▄·█·▄·▄ beats 3/5	very hard · avg 4.3%	▇▇▇ 6.4%	several runs
GPQA (subset unspecified) lab +1	79.1% ✎	▂▁▂▂▄▃▆▃▅█▄▃ beats 93/104	hard · avg 47%	▇▇ 4.6%	sharp · at its level · small run · several runs
MMLU-Pro +1	81.5% ✎	▄·▄···█·▄··▄ beats 0/5	medium · avg 56%	▇▇ 4.2%	several runs
MMLU-Redux	89.9% ✎	████··█····█ beats 1/5	medium · avg 73%	▇ 2.5%	typical
EQ-Bench 3 (emotional intelligence, judge Elo)	1229.2 ✎	▁·▂▁▄▅▅▃█▄▁▁ beats 36/62	avg 1044	▇ 2.1%	typical
ZebraLogic	90.9% ✎	██··█·██···█ beats 0/5	easy · avg 75%	▇ 2.0%	maxed out
MMLU lab	90.0% ✎	▁▁▁▁▁▂▃▅█▃▆▃ beats 179/183	medium · avg 72%	▇ 1.6%	small run

9 more tests

test	result	vs the others in its run	difficulty	counts	why
VitaBench	26.8% ✎	█····██·██·█ beats 3/5	very hard · avg 7.8%	▇ 1.5%	typical
OJBench	19.0% ✎	▄······▄▄·▄█ beats 0/5	very hard · avg 12%	▇ 1.5%	at its level
RPC-Bench overall, text input (Z.ai) lab	25.7% ✎	▄··▄▄·▄▄█·▄▄ beats 2/8	very hard · avg 11%	▇ 1.3%	typical
SimpleQA lab	26.4% ✎	▄▄▄▆█▆███·▂█ beats 25/32	very hard · avg 7.7%	▇ 1.2%	small run
BeyondAIME	66.0% ✎	█··▄····▄··█ beats 2/5	medium · avg 41%	▇ 1.1%	at its level
MultiChallenge lab	52.8% ✎	▁·▁·▁▃▆▅▃▁█▅ beats 16/29	hard · avg 27%	▇ 0.9%	at its level · small run
miniF2F (pass@1)	10.9% ✎	██······▄··▄ beats 0/5	very hard · avg 13%	▇ 0.1%	blunt · at its level · noisy
Vectara hallucination leaderboard, 2025 dataset (HHEM-2.3, factual consistency rate)	96.3% ✎	▁▁▂▃▄▄█▆▃▂▂▁ beats 93/140	very easy · avg 94%	▇ 0.1%	blunt · maxed out · noisy
SysBench (ISR) lab	81.0% ✎	█··········█ beats 1/1		—	not counted yet: no other model in the same run to compare it with

$ where to run it

Through an API

api · $1.00/M · $0.6 in / $2.2 out per 1M tokens · 131k context · docs.z.ai ↗ · price sourceStandard base token rate from models.dev. Provider-specific tiers, long-context surcharges, caching and promotions may differ.

llama.cpp · recommended

version

Install llama.cpp · Homebrew and winget installs name the commands llama-cli and llama-server instead.
macOS / Linux
curl -LsSf https://llama.app/install.sh | sh
Windows
irm https://llama.app/install.ps1 | iex
other ways to install ↗
Run GLM-4.5 · needs ~213.5 GB · from huggingface.co/unsloth
chat in the terminal
llama cli -hf unsloth/GLM-4.5-GGUF:UD-Q4_K_XL
Or start a local server and open the address it prints (http://localhost:9931)
llama serve -hf unsloth/GLM-4.5-GGUF:UD-Q4_K_XL

LM Studio

version

Install LM Studio · Open the app once before using the lms command.
download it (macOS / Windows / Linux) ↗
Run GLM-4.5 · needs ~213.5 GB · from huggingface.co/unsloth
open in LM Studio ↗
Or download it in the terminal
lms get https://huggingface.co/unsloth/GLM-4.5-GGUF@UD-Q4_K_XL -y
Then run lms ls to see its name, and lms chat followed by that name

MLX (Mac)

Install MLX (Mac) · Apple silicon Macs only, text only.
with uv
uv tool install mlx-lm
or
pip install mlx-lm
Run GLM-4.5 · needs ~198.6 GB, Mac only · from huggingface.co/mlx-community
chat in the terminal
mlx_lm.chat --model mlx-community/GLM-4.5-4bit
Or start an OpenAI-compatible server at http://localhost:8080/v1
mlx_lm.server --model mlx-community/GLM-4.5-4bit

Weights and documentation: huggingface.co/zai-org ↗

$ how the score was worked out · Meituan LongCat-Flash-Thinkin… 41%, GLM-4.5 Team (Zhipu AI & Tsin… 21%, Prime Intellect INTELLECT-3 m… 19% and 6 more

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The 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's added up by who ran the tests. How it works

source	results	share of the score	licence
Meituan LongCat-Flash-Thinking model card (evaluation table)	18	▇▇▇▇▇▇▇▇▇▇▇▇ 41%	MIT
GLM-4.5 Team (Zhipu AI & Tsinghua) — GLM-4.5 technical report · lab-reported	16	▇▇▇▇▇▇ 21%	CC BY 4.0
Prime Intellect INTELLECT-3 model card: Evaluation table (Environments Hub runs)	6	▇▇▇▇▇▇ 19%	MIT
zai-org/GLM-4.6 model card · lab-reported	7	▇▇▇ 11%	MIT
LMArena leaderboard dataset	1	▇ 3%	CC BY 4.0
EQ-Bench 3 canonical leaderboard results	1	▇ 2%	MIT
RPC-Bench leaderboard (Z.ai research paper comprehension) · lab-reported	1	▇ 1%	Apache-2.0
Berkeley Function Calling Leaderboard result archive	1	▇ 1%	Apache-2.0
Vectara hallucination leaderboard (vectara/hallucination-leaderboard repository)	1	▇ 0%	Apache-2.0
prior · pull toward the middle		▇ 0%
weighted average		= 64.7

not counted (3)

GLM-4.5 Team (Zhipu AI & Tsinghua) — GLM-4.5 technical report	LiveCodeBench 2024-07 to 2025-01 (version not stated) [correct]	72.9%	fewer than three models have this test yet, or this source ran it for this model alone
GLM-4.5 Team (Zhipu AI & Tsinghua) — GLM-4.5 technical report	SysBench (ISR) [correct]	81.0%	fewer than three models have this test yet, or this source ran it for this model alone
zai-org/GLM-4.6 model card	τ²-bench overall (weighted) [correct]	67.5%	fewer than three models have this test yet, or this source ran it for this model alone

$ improve this page

Update model information Add a place to run it Add a test result Propose a benchmark Correct a benchmark score Edit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated 2026-10-11 02:20 UTC · sources and licences · json