GLM-4.6V

+ contribute

GLM-4.6V is an open-weights language model from Z.ai, released December 2025. Its published size is 106B parameters and a 128k-token context window. Licence: mit.

Catalogue identity notes

Arena entry 'glm-4.6v' (text, vision). Maker repo zai-org/GLM-4.6V @4e2d47e: licence tag mit. Parameters: card: 'GLM-4.6V (106B), a foundation model designed for cloud ...' (active count not stated on the card). Released 2025-12: month the maker's Hugging Face repo was created (createdAt 2025-12-07).

62.2±1.7#175 of 719 LLMsfrom 10 benchmark results across 3 sources

$ how it compares

The LLMs ranked either side of GLM-4.6V, out of 828 with a score, then more from Z.ai.

#	model	score	where	RAM	released
203	GPT-4.1 OpenAI	62±1	api $3.50/M		2025-04
204	Nemotron 3.5 Lightning 30B-A3B NVIDIA	62±2	api $0.087/M +local	~20GB	2026-08
205	GLM-4.6V Z.ai	62±2	api $0.45/M +local	~51GB	2025-12
206	Devstral 2 (2512) Mistral	62±2	local	~75GB	2025-12
207	Gemini 2.5 Flash Preview (05-20) Google	62±1	comparison only		2025-05
	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

chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #129 of 393 * · 27% of its score · 1 test

Beats LongCat 2.0 (1.6T-A48B), o3-mini, K-EXAONE (236B-A23B) here, though they rank higher overall. Trails LFM2.5 8B-A1B, LFM2.5 2.6B, Nemotron-Cascade-8B here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
LMArena (text)	1378.7 ✎	▁▁▁▂▃▃▃▆▆▇█▃ beats 219/379	avg 1268	▇▇▇▇▇▇▇▇▇▇▇ 27%	blunt

vision & documents▁▁▃▂▁▃▄▃▆▇█▄ #88 of 147 · 61% of its score · 7 tests

Beats DeepSeek V4 Flash Vision (Exp) here, though they rank higher overall. Trails Mistral Small 4 (119B-A6.5B), Qwen3-VL 32B Instruct, GPT-4.1 mini here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Arena vision (style controlled)	1161.5 ✎	▁▁▂▂▂▂▄▅▅▆██ beats 56/145	avg 1068	▇▇▇▇▇▇▇ 18%	blunt
CharXiv (reasoning questions) lab	63.2% ✎	▂▅▃▅▅▆▆▅▆█▆▂ beats 48/100	medium · avg 38%	▇▇▇▇ 10%	at its level · small run
SciDocBench scientific-document workflows (overall, 496 instances)	42.4% ✎	▃▃···█▃▅▃·▃▅ beats 1/11	hard · avg 26%	▇▇▇ 8.5%	blunt · at its level
MMMU (val) lab	76.0% ✎	▁····▁▁▄█▇▆▆ beats 41/52	medium · avg 58%	▇▇▇ 8.5%	at its level · small run
MMMU Pro lab	66.0% ✎	▂▃▁▃▆▇▃▇▇██▅ beats 26/52	medium · avg 46%	▇▇▇ 8.0%	at its level · small run
MMStar lab	75.9% ✎	▃▅▂▃▂▂▆█▆▆█▅ beats 22/34	medium · avg 63%	▇▇▇ 6.3%	at its level · small run
OCRBench lab	86.5% ✎	▁·▃▁▆▅▂█▅█▁▃ beats 21/37	easy · avg 83%	▇ 1.9%	blunt · small run

other tests2 not in a task · 11% of its score

test	result	vs the others in its run	difficulty	counts	why
LVBench lab	59.5% ✎	▇·█▅▅▄▄··▃▁▁ beats 12/28	hard · avg 49%	▇▇ 5.7%	at its level · small run
AndroidWorld lab	57.0% ✎	▁·▁▁▂▃█▅▂▂·▂ beats 11/24	medium · avg 41%	▇▇ 5.1%	blunt · at its level · small run

$ where to run it

Through an API

api · $0.45/M · $0.3 in / $0.9 out per 1M tokens · 128k context · docs.z.ai ↗ · price sourceUSD per 1M tokens, Z.ai API pricing page, Vision Models table, GLM-4.6V: input $0.3, output $0.9; cached input $0.05 excluded. Currency: USD. Z.ai official pricing page, observed 2026-09-28.

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-4.6V · needs ~67.22 GB · from huggingface.co/unsloth
chat in the terminal
llama cli -hf unsloth/GLM-4.6V-GGUF:UD-Q4_K_XL
Or start a local server and open the address it prints (http://localhost:9931)
llama serve -hf unsloth/GLM-4.6V-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-4.6V · needs ~67.22 GB · from huggingface.co/unsloth
open in LM Studio ↗
Or download it in the terminal
lms get https://huggingface.co/unsloth/GLM-4.6V-GGUF@UD-Q4_K_XL -y
Then run lms ls to see its name, and lms chat followed by that name

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

$ how the score was worked out · LMArena leaderboard dataset 45%, Z.ai GLM-V GitHub repository… 45%, SciDocBench: A Workflow-Cente… 9%

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
LMArena leaderboard dataset	2	▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 45%	CC BY 4.0
Z.ai GLM-V GitHub repository benchmark image (GLM-4.6V) · lab-reported	7	▇▇▇▇▇▇▇▇▇▇▇▇▇ 45%	Apache-2.0
SciDocBench: A Workflow-Centered Benchmark and Data Pipeline for Scientific Document Understanding (arXiv:2609.05141v2)	1	▇▇▇ 9%	CC BY 4.0
prior · pull toward the middle		▇ 1%
weighted average		= 62.2

$ 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