GLM-5.1

zai-org MIT out 7 months ago

754B parameters (8 of 256 experts per token), text in, text out, 198K tokens of context. 15 providers serve it from $0.96 per million input tokens.

Verified 2026-10-25 Rank #52 in LLMs · score 107.4 method v1 Sources 4 Downloads 30d 702.2k · #139 of 619 tracked

Can I run it?452GB at INT4Estimateneeds a multi-GPU nodemethodCommunity: I ran GLM-5.1 on a 16GB RAM machine →

What does it cost?$0.96/1M incheapest of 12Baidufp82026-10-24

Can I use it?MITPermissivelicence file2026-10-24

Do people use it?702.2kdownloads, 30 days#139 of 619 we track · top 22%+4 likes, downloads +29% this week1,850 likesHugging Face2026-10-24

Worth reading

Picked by rule from 22 tracked pieces: the lab's own post, the most shared, a field report, the latest.

Off-peak usage for GLM-5.1 is billed at 1× as a limited-time promotion through the end of April.

From GLM-5.1: Towards Long-Horizon Tasks · z.ai · 2026-04-07

The lab's own post

GLM-5.1: Towards Long-Horizon Tasks

GLM-5.1 sustains optimization over hundreds of rounds and thousands of tool calls by revisiting reasoning and revising strategy through repeated iteration.

The model temporarily breaks the Recall constraint during exploration before adjusting to restore it after major structural transitions.
Off-peak usage for GLM-5.1 is billed at 1× as a limited-time promotion through the end of April.

z.ai 2026-04-07 618 HN pts · 263 comments 2.5k words

Most discussed · 32 HN pts

GLM-5.1 Is Available

twitter.com 2026-03-27 32 HN pts · 6 comments

Latest · 5 months ago

I ran GLM-5.1 on a 16GB RAM machine

github.com 2026-05-27 10 HN pts · 1 comments

22 best of 22 tracked

The 4 picks above, plus these 18. Top is the radar's score: discussion, source authority and the lab's own posts, fading with age.

ArticleGLM 5.1docs.z.ai · 2026-03-27 · 6 HN pts · official
ArticleGLM 5.1: Pelican Testsimonwillison.net · 2026-04-08 · 4 HN pts
ArticleGLM-5.1 (Reasoning) vs Claude Opus 4.6 (Non- ...artificialanalysis.ai · undated
DiscussionGLM-5.1 Releasedtwitter.com · 2026-03-27 · 11 HN pts
ArticleComparing GPT-5.4, Opus 4.6, GLM-5.1, Kimi K2.5, MiMo V2 Pro and MiniMax M2.7codejam.info · 2026-04-18 · 6 HN pts
ArticleGLM 5.1 Review: A Preview Model Tested Hard | Skalablogskalablog.com · 2026-10-01
DiscussionMy GLM-5.1 coding agent scored 94.3% on LiveCodeBench Lite (348/369)news.ycombinator.com · 2026-04-22
DiscussionGLM-5.1twitter.com · 2026-04-07
ArticleGLM-5.1 tops Vector DB Benchmarkvector-db-bench.kcores.com · 2026-04-01
DiscussionGLM 5.1 Is Outold.reddit.com · 2026-03-27
ArticleGLM-5.1 Deep Dive: Open-Source MoE Model for Agentic Codingcodexpedite.com · 2026-05-19
ArticleGLM-5.1 — Open-Weight Frontier from the World's First Publicly Listed LLM Company — ChatForestchatforest.com · 2026-05-13
ArticleGLM-5.1: 754B MoE model, AIME 95.3%, SWE-Bench Pro 58.4%, MIT licensed — Reebootreeboot.fr · 2026-05-13
ArticleGLM-5.1 & Zhipu AIcodexpedite.com · 2026-05-09
ArticleGLM-5.1: Architecture, Benchmarks, Capabilities & How to ...analyticsvidhya.com · undated
ArticleGLM-5.1 Model Overview: Features, Capabilities & Use Casesdeepinfra.com · undated
ArticleGLM-5.1 · LLM Model Review | CrucibleMarkcruciblemark.com · undated
ArticleGLM-5.1 Review: Open-Source Model Tops SWE-Bench Pro | Awesome Agentsawesomeagents.ai · 2026-04-17

Show all 18

Where to run it

Live hosted endpoints, cheapest first, USD per million tokens. Ties break on output price, then provider name.

12 providers from $0.96

12 providers with a live price, from $0.96 to $1.40 per million input tokens; the median is $1.23. One dot per provider, cheapest in teal.

#	Cheapest 5	In /1M	Out /1M	Context	Quant	Seen
1	Baidu	$0.96	$3.03	198K	fp8	Oct 24
2	StreamLake	$0.97	$3.04	195K	fp8	Oct 24
3	Chutes	$0.98	$3.08	198K	fp8	Oct 24
4	DeepInfra	$1.05	$3.50	198K	—	Oct 25
5	SiliconFlow	$1.19	$3.74	200K	fp8	Oct 24

All 15 providers →

Where it's used

Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.

Widely served (15 API providers); 33 quantized builds for local use.

As a service15Alibaba, AtlasCloud, Baidu, Chutes +11
On own hardware33quantized builds
Built on1211 fine-tunes · 1 adapters · 0 merges
In apps84Hugging Face Spaces

Top Spaces running it: quickgrid/Tokenizer-Visualizer ♥ 6 · lujin/zai-org-GLM-5.1 ♥ 4 · victor/front-skill-bakeoff ♥ 3 · lvwerra/self-plan ♥ 2

Will it fit?

Weights only, an estimate: leave room for the KV cache. Method

Needs an 8×80 GB node even at INT4

About 452 GB at INT4, 905 GB at FP8, 1.81 TB at BF16: weights only, so leave room for the KV cache. Every expert must be resident, so total parameters count.

Estimate params × bytes per param × 1.2 · how we estimate

The story so far

Weights, every dated article, and repo commits on one rail.

Coverage peaked at launch and is still arriving

Teal is the weights (2026-04-03), amber triangles are 9 dated articles (each links to it), grey ticks are repo commits.

What people built on it

45 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.

45 derivatives on the Hub, most of them quantized

From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.

GGUF unsloth/GLM-5.1-GGUF for llama.cpp, Ollama, LM Studio 250.8k downloads MLX baa-ai/GLM-5.1-RAM-420GB-MLX for Apple silicon 12k downloads NVFP4 nvidia/GLM-5.1-NVFP4 for vLLM, SGLang 178.9k downloads

All variants →

Reference

Everything else, folded: the full tables, every source.

Benchmarks

28 rows

Benchmarks 28 rows

measured third-party leaderboard rowauthor-reported number from the model card or eval files (28)No confidence intervals unless the source gave them; settings differ between rows, so compare within a benchmark, not across. Date = the benchmark's own date when it gives one, else when we read it.

Benchmark	Value	Setting / notes	Reported by	Date
author-reported AIME 2026 · Reason through hard problems	95.3	—	model author	2026-10-07
author-reported BrowseComp	68	—	model author	2026-10-07
author-reported BrowseComp (w/ Context Manage)	79.3	—	model author	2026-10-07
author-reported CyberGym	68.7	—	model author	2026-10-07
author-reported GPQA Diamond · Reason through hard problems	86.2	—	model author	2026-10-07
author-reported HLE (w/ Tools)	52.3	—	model author	2026-10-07
author-reported HMMT · Reason through hard problems	94	—	model author	2026-10-07
author-reported HMMT · Reason through hard problems	82.6	—	model author	2026-10-07
author-reported Humanity's Last Exam · Reason through hard problems	31	—	model author	2026-10-07
author-reported IMO-AnswerBench · Reason through hard problems	83.8	—	model author	2026-10-07
author-reported MCP-Atlas (Public Set)	71.8	—	model author	2026-10-07
author-reported NL2Repo · Write code with agents	42.7	—	model author	2026-10-07
author-reported SWE-bench Pro · Write code with agents	58.4	—	model author	2026-10-07
author-reported Terminal-Bench · Write code with agents	63.5	—	model author	2026-10-07
author-reported Terminal-Bench · Write code with agents	69	Claude Code	model author	2026-10-07
author-reported Tool-Decathlon	40.7	—	model author	2026-10-07
author-reported Vending Bench 2	$5,634.41 other	—	model author	2026-10-07
author-reported τ³-Bench	70.6	—	model author	2026-10-07
author-reported MathArena/aime_2026	95.3	—	Model Card	2026-04-07
author-reported MathArena/hmmt_feb_2026	82.6	—	Model Card	2026-04-07
author-reported general (claw-eval/Claw-Eval)	62.7	Pass³% \| N=3 \| 161 tasks	Claw-Eval Leaderboard	2026-04-23
author-reported multi_turn (claw-eval/Claw-Eval)	60.5	Pass³% \| N=3 \| 38 tasks	Claw-Eval Leaderboard	2026-04-23
author-reported diamond (Idavidrein/gpqa)	86.2	—	Model Card	2026-04-07
author-reported hle (cais/hle)	31	—	Model Card	2026-04-07
author-reported hle (cais/hle)	52.3	With tools	Model Card	2026-04-07

Show all 28 rows (3 more)

Benchmark	Value	Setting / notes	Reported by	Date
author-reported SWE_Bench_Pro (ScaleAI/SWE-bench_Pro)	58.4	high reasoning	Model Card	2026-10-24
author-reported terminalbench_2 (harborframework/terminal-bench-2.0)	63.5	—	Model Card	2026-04-07
author-reported terminalbench_2 (harborframework/terminal-bench-2.0)	69	agent: Terminus 2(Claude Code)	Model Card	2026-04-07

Use-case evidence

5 use cases

What it is good at 5 use cases

Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.

Facts

24 fields

Listed in LLMs · MoE

Facts

GLM-5.1 has 754B parameters. (safetensors metadata, observed 2026-10-24)
GLM-5.1 is a mixture-of-experts model with 256 experts, routing 8 per token. (config.json, observed 2026-10-24)
Context length is 202,752 tokens (198K). (config.json max_position_embeddings, observed 2026-10-24)
License: MIT (friction: none). Permissive. (license file, observed 2026-10-24)
Created 2026-04-03, last modified 2026-05-13. (HF API, observed 2026-10-24)

Pipeline tag	text-generation
Library	transformers
Architectures	GlmMoeDsaForCausalLM
Model type	glm_moe_dsa
Dtype	bfloat16
Params by dtype	BF16: 753,864,119,552 · F32: 19,456
Size band	XXL
MoE	yes
Experts	256
Experts per token	8
Context length	202,752 (config.json max_position_embeddings)
Languages	en, zh
Papers	arXiv:2602.15763
License	MIT (mit)
Hub snapshot	2026-10-25

Servability

vLLM · SGLang

Servability

vLLM: supported · GlmMoeDsaForCausalLM · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: not in registry · 773cb938fe9484a8f7e7e46fb1b1fe96412c59ed

vLLM registry (7050ef881ff0835f6890a47db1859a24c65481d6) · SGLang registry (773cb938fe9484a8f7e7e46fb1b1fe96412c59ed) · Matched by architecture class name against each project's registry on main; 'supported' means the architecture is registered, not that this checkpoint was tested.

Hardware fit

3 estimates

Hardware fit ESTIMATE

Weights-only memory for GLM-5.1: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.

BF16 (2 bytes/param)	1.81 TB · Cluster (≥640 GB)
FP8 (1 byte/param)	905 GB · Cluster (≥640 GB)
INT4 (0.5 bytes/param)	452 GB · Multi-GPU node (<640 GB)

Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra. MoE: all experts must be resident, so total parameters count.

All variants

40 variants

Lineage

Original model Family: zai-org/GLM-5.1 45 derivatives on the Hub

quantized: 33finetune: 11adapter: 1

Top variants (40 shown, ranked by downloads)

Variant	Relation	Quant	Params	Band	License	Author	Likes	Downloads
unsloth/GLM-5.1-GGUF	quantized	GGUF	—	—	mit	unsloth	205	250.8k
nvidia/GLM-5.1-NVFP4	quantized	NVFP4	382B	XXL	mit	nvidia	46	178.9k
lukealonso/GLM-5.1-NVFP4	quantized	NVFP4	437B	XXL	mit	lukealonso	23	160.8k
amd/GLM-5.1-MXFP4	quantized	MXFP4	408B	XXL	mit	amd	1	49.4k
bartowski/zai-org_GLM-5.1-GGUF	quantized	GGUF	—	—	mit	bartowski	12	31.8k
cyankiwi/GLM-5.1-AWQ-4bit	quantized	AWQ	744B	XXL	mit	cyankiwi	13	25.5k
ubergarm/GLM-5.1-GGUF	quantized	GGUF	—	—	mit	ubergarm	26	12.8k
baa-ai/GLM-5.1-RAM-420GB-MLX	quantized	MLX	744B	XXL	other	baa-ai	5	12k
mlx-community/GLM-5.1-DQ4plus-q8	quantized	—	744B	XXL	mit	mlx-community	6	7.7k
catalystsec/GLM-5.1-4bit	quantized	-4bit	744B	XXL	mit	catalystsec	2	7.2k
mindlab-research/Macaron-V1-Preview-749B	finetune	—	754B	XXL	mit	mindlab-research	83	2.8k
spicyneuron/GLM-5.1-MLX-2.9bit	quantized	MLX	744B	XXL	mit	spicyneuron	2	6.7k
0xSero/GLM-5.1-478B-NVFP4	quantized	NVFP4	280B	XXL	mit	0xSero	10	5.9k
baa-ai/GLM-5.1-RAM-270GB-MLX	quantized	MLX	744B	XXL	other	baa-ai	2	6.1k
sokann/GLM-5.1-GGUF-1.673bpw	quantized	GGUF	—	—	mit	sokann	5	5.5k
DevQuasar/zai-org.GLM-5.1-GGUF	quantized	GGUF	—	—	—	DevQuasar	4	5.5k
storagejuju/glm-5.1-ud-iq2-m-juju	finetune	—	—	—	mit	storagejuju	0	5.6k
Nishant2414/GLM-5.1-MLX-4.8bit	quantized	MLX	—	—	—	Nishant2414	1	3.4k
spicyneuron/GLM-5.1-MLX-3.6bit	quantized	MLX	744B	XXL	mit	spicyneuron	1	2.8k
0xSero/GLM-5.1-555B-GGUF	quantized	GGUF	—	—	other	0xSero	7	2.5k
spicyneuron/GLM-5.1-MLX-4.5bit	quantized	MLX	744B	XXL	mit	spicyneuron	1	2.5k
sombra-x/GLM-5.1-REAP-Solidity-MXFP8	quantized	—	381B	XXL	mit	sombra-x	0	2.3k
0xSero/GLM-5.1-555B	finetune	—	563B	XXL	other	0xSero	2	2k
unsloth/GLM-5.1-FP8	quantized	-FP8	754B	XXL	mit	unsloth	5	1.8k
tiny-random/glm-5.1	finetune	—	8.70M	S	—	tiny-random	1	1.8k
koushd/GLM-5.1-NVFP4	quantized	NVFP4	436B	XXL	mit	koushd	1	1.7k
livadies/GLM-5.1-Ghetto-MoE-2-Experts	finetune	—	2.70B	M	—	livadies	4	1.5k
QuantPasture/GLM-5.1-GGUF	quantized	GGUF	—	—	—	QuantPasture	2	1.4k
yujiepan/glm-5.1-tiny-random	finetune	—	8.70M	S	—	yujiepan	0	1.4k
mconcat/GLM-5.1-FP8-Dynamic	quantized	-FP8	744B	XXL	mit	mconcat	0	1.2k
0xSero/GLM-5.1-444B	finetune	—	455B	XXL	other	0xSero	2	963
0xSero/GLM-5.1-444B-GGUF	quantized	GGUF	—	—	other	0xSero	1	810
mindlab-research/Macaron-V1-Preview-744B-Merged	finetune	—	754B	XXL	mit	mindlab-research	5	477
amd/GLM-5.1-NVFP4	quantized	NVFP4	380B	XXL	mit	amd	0	499
jybsuper/GLM-5.1-6layer	finetune	—	42.7B	XL	other	jybsuper	0	318
Kakarikukuru/Huihui-GLM-5.1-abliterated-GGUF	quantized	GGUF	—	—	mit	Kakarikukuru	1	222
andreasnbg/Huihui-GLM-5.1-abliterated-GGUF	quantized	GGUF	—	—	mit	andreasnbg	0	246
mindlab-research/Macaron-A2UI-Venti	adapter	—	—	—	mit	mindlab-research	2	80
CPU-Hybrid-MoE/GLM-5.1-CPU-NUMA4-AMXINT8	finetune	—	—	—	—	CPU-Hybrid-MoE	1	121
rbinrs/Huihui-GLM-5.1-abliterated-GGUF	quantized	GGUF	—	—	mit	rbinrs	2	71

Repository changelog

17 commits

Repository changelog 17 commits

Date	Commit	SHA
2026-05-13	Add Claw-Eval evaluation results (#33)	26e1bd6
2026-04-16	update chat_template	442b527
2026-04-12	update CyberGym score for Gemini 3.1 Pro and GPT-5.4	bef0214
2026-04-11	Merge branch 'main' of hf.co:zai-org/GLM-5.1	0e12244
2026-04-11	update	7015f16
2026-04-08	Update .eval_results/swe_bench_pro.yaml (#10)	dd43008
2026-04-08	update bench	146e9f2
2026-04-07	add SWE-bench_Pro	278f446
2026-04-07	GLM-5.1	bf3dec1
2026-04-07	Update README.md	ada0c51
2026-04-04	Add files using upload-large-folder tool	56bbdce
2026-04-04	Add files using upload-large-folder tool	5b48738
2026-04-04	Add files using upload-large-folder tool	6f7b6cb
2026-04-04	Add files using upload-large-folder tool	12ec8fb
2026-04-04	Add files using upload-large-folder tool	a1ec45e
2026-04-04	Add files using upload-large-folder tool	f5ed80e
2026-04-03	initial commit	385269b

All providers

15 providers · 15 endpoints

Where to run 15 providers · 15 endpoints

Hosted endpoints observed for GLM-5.1, sorted by input price (USD per million tokens). Each row names its observer and date; prices change daily, so check the source before quoting. Cheapest observed input price: $0.96/M.

Provider	Model ref	Context	Quant	$/M in	$/M out	tok/s	TTFT	Tools	Status	Observed
Baidu openrouter	z-ai/glm-5.1	198K	fp8	$0.96	$3.03	—	—	yes	live	2026-10-24
StreamLake openrouter	z-ai/glm-5.1	195K	fp8	$0.97	$3.04	—	—	yes	live	2026-10-24
Chutes openrouter	z-ai/glm-5.1	198K	fp8	$0.98	$3.08	—	—	yes	live	2026-10-24
DeepInfra hf	zai-org/GLM-5.1	198K	—	$1.05	$3.50	74.8	500 ms	yes	live	2026-10-25
SiliconFlow openrouter	z-ai/glm-5.1	200K	fp8	$1.19	$3.74	—	—	yes	live	2026-10-24
Phala openrouter	z-ai/glm-5.1	198K	unknown	$1.21	$4.20	—	—	yes	live	2026-10-24
AtlasCloud openrouter	z-ai/glm-5.1	198K	fp8	$1.26	$3.96	—	—	yes	live	2026-10-24
Alibaba openrouter	z-ai/glm-5.1	198K	fp8	$1.33	$4.18	—	—	yes	live	2026-10-24
Novita hf, openrouter	zai-org/glm-5.1	200K	fp8	$1.38	$4.40	47.8	2681 ms	yes	staging	2026-10-25
Nebius openrouter	z-ai/glm-5.1	198K	fp8	$1.40	$4.40	—	—	yes	status -5	2026-10-24
GMICloud openrouter	z-ai/glm-5.1	198K	fp8	$1.40	$4.40	—	—	—	live	2026-10-24
Friendli openrouter	z-ai/glm-5.1	198K	unknown	$1.40	$4.40	—	—	yes	live	2026-10-24
Z.AI openrouter	z-ai/glm-5.1	198K	fp8	$1.40	$4.40	—	—	yes	live	2026-10-24
Venice openrouter	z-ai/glm-5.1	195K	fp8	$1.40	$4.40	—	—	yes	live	2026-10-24
Featherless hf	zai-org/GLM-5.1	—	—	—	—	—	—	—	live	2026-10-25

Articles and coverage

22 pieces

Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency. Summaries are generated from a stored copy of the article text; every sentence cites a span of that text, and nothing is shown that failed the citation check.

Official2

GLM-5.1: Towards Long-Horizon Tasks

z.ai2026-04-07officialauthorityHN front pageHN 618 pts · 263 comments2,489 words

GLM-5.1 sustains optimization over hundreds of rounds and thousands of tool calls by revisiting reasoning and revising strategy through repeated iteration. Unlike previous models that plateau early, it achieves structural transitions like IVF cluster probing and two-stage pipelines to shift performance frontiers. The model reaches 21.5k QPS after 600+ iterations and 6,000+ tool calls on vector database optimization tasks.written, reviewed, cited

The model temporarily breaks the Recall constraint during exploration before adjusting to restore it after major structural transitions.
Off-peak usage for GLM-5.1 is billed at 1× as a limited-time promotion through the end of April.

GLM 5.1

docs.z.ai2026-03-27officialauthorityHN 6 pts · 0 comments1,257 words

The GLM Coding Plan now supports the latest GLM-5.1 model for all users (Max, Pro, and Lite), and you can switch between models within your preferred Coding Agent. Users edit ~/.claude/settings.json or ~/.openclaw/openclaw.json to add the model, setting the Context Window Size to 200000. This currently only works with coding agents that support custom model configuration.written, reviewed, cited

Users must uncheck the Support Images option when configuring the custom GLM-5.1 model.
Engineers using OpenClaw must manually update the configuration after completing the provider setup.
Agents lacking custom model settings require waiting for official support in a future release.

Analysis6

I ran GLM-5.1 on a 16GB RAM machine

github.com2026-05-27authorityHN 10 pts · 1 comments

GLM 5.1: Pelican Test

simonwillison.net2026-04-08authorityHN 4 pts · 2 comments394 words

Z.ai released GLM-5.1, a 754B parameter 1.51TB MIT-licensed model available via OpenRouter. The model autonomously generates complex HTML+SVG outputs with CSS animations for long-horizon tasks. It demonstrates reasoning capabilities by identifying and fixing positioning conflicts between CSS transforms and SVG attributes.written, reviewed, cited

GLM-5.1 matches Opus 4.6 in agentic performance, at ~1/3 actual cost

app.uniclaw.ai2026-04-07HN 22 pts · 2 comments69 words

GLM-5.1 (Reasoning) vs Claude Opus 4.6 (Non- ...

artificialanalysis.aiauthority816 words

Comparing GPT-5.4, Opus 4.6, GLM-5.1, Kimi K2.5, MiMo V2 Pro and MiniMax M2.7

codejam.info2026-04-18HN 6 pts · 2 comments1,457 words

GPT-5.4 achieved an average cleanliness score of 9 while GLM-5.1 scored 8 in the technical approach evaluation. Kimi K2.5 and MiniMax M2.7 were consistently ranked at the bottom with average scores of 5.6 and 5.2 respectively.generated, cited

GLM 5.1 Review: A Preview Model Tested Hard | Skalablog

skalablog.com2026-10-01

Community2

GLM-5.1 Is Available

twitter.com2026-03-27HN 32 pts · 6 comments

GLM-5.1 Released

twitter.com2026-03-27HN 11 pts · 2 comments

All coverage tracked for this model (22) — append-only, nothing is ever removed

Title	Source	Published	First seen	Signals
GLM-5.1 (Reasoning) vs Claude Opus 4.6 (Non- ...	artificialanalysis.ai	—	2026-10-07	exa
GLM-5.1: Architecture, Benchmarks, Capabilities & How to ...	analyticsvidhya.com	—	2026-10-07	exa
GLM-5.1 Model Overview: Features, Capabilities & Use Cases	deepinfra.com	—	2026-10-07	exa
GLM-5.1 · LLM Model Review \| CrucibleMark	cruciblemark.com	—	2026-10-07	exa
GLM 5.1 Review: A Preview Model Tested Hard \| Skalablog	skalablog.com	2026-10-01	2026-10-07	exa
I ran GLM-5.1 on a 16GB RAM machine	github.com	2026-05-27	2026-10-07	hn
GLM-5.1 Deep Dive: Open-Source MoE Model for Agentic Coding	codexpedite.com	2026-05-19	2026-10-07	exa
GLM-5.1 — Open-Weight Frontier from the World's First Publicly Listed LLM Company — ChatForest	chatforest.com	2026-05-13	2026-10-07	exa
GLM-5.1: 754B MoE model, AIME 95.3%, SWE-Bench Pro 58.4%, MIT licensed — Reeboot	reeboot.fr	2026-05-13	2026-10-07	exa
GLM-5.1 & Zhipu AI	codexpedite.com	2026-05-09	2026-10-07	exa
My GLM-5.1 coding agent scored 94.3% on LiveCodeBench Lite (348/369)	news.ycombinator.com	2026-04-22	2026-10-07	hn
Comparing GPT-5.4, Opus 4.6, GLM-5.1, Kimi K2.5, MiMo V2 Pro and MiniMax M2.7	codejam.info	2026-04-18	2026-10-07	hn
GLM-5.1 Review: Open-Source Model Tops SWE-Bench Pro \| Awesome Agents	awesomeagents.ai	2026-04-17	2026-10-07	exa
GLM 5.1: Pelican Test	simonwillison.net	2026-04-08	2026-10-07	hn
GLM-5.1 matches Opus 4.6 in agentic performance, at ~1/3 actual cost	app.uniclaw.ai	2026-04-07	2026-10-07	hn
GLM-5.1	twitter.com	2026-04-07	2026-10-07	hn
GLM-5.1: Towards Long-Horizon Tasks	z.ai official	2026-04-07	2026-10-07	hn
GLM-5.1 tops Vector DB Benchmark	vector-db-bench.kcores.com	2026-04-01	2026-10-07	hn
GLM 5.1 Is Out	old.reddit.com	2026-03-27	2026-10-07	hn
GLM 5.1	docs.z.ai official	2026-03-27	2026-10-07	hn
GLM-5.1 Released	twitter.com	2026-03-27	2026-10-07	hn
GLM-5.1 Is Available	twitter.com	2026-03-27	2026-10-07	hn

Papers

1 paper

Papers 1

arXiv links found in the model card and its tags. "Peer-reviewed" is shown only with evidence: a published venue in OpenAlex (measured) or the authors' own journal reference on arXiv (reported).

GLM-5: from Vibe Coding to Agentic Engineering this model

GLM-5-Team, Aohan Zeng, Xin Lv +183 · 2026-02-17 · PDF · HF paper page (223 upvotes)

Similar models

3 models

Similar models

GLM-5.3 · 753B
same size band (XXL), also MoE, same task (text-generation), more providers
GLM-5.2 · 753B
same size band (XXL), also MoE, same task (text-generation), more providers, permissive license
GLM-5 · 754B
same size band (XXL), also MoE, same task (text-generation), more providers, permissive license

Sources

4 sources

Sources

Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:33 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:45 UTC · Summary: 2026-10-07 15:00 UTC · Last verified 2026-10-25

Follow GLM-5.1

Watch this model

/feeds/models/zai-org--GLM-5.1.xml(RSS 2.0) · every change to GLM-5.1
/feeds/families/zai-org--GLM-5.1.xml(RSS 2.0) · every change across the GLM-5.1 family

Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.