GLM 5.1

Apr 2026
GLM 5.1
Compare in
on
$1.02per task?
$1.40 / $4.40per 1M in / out?cache $0.26
72 tok/sfirst token in 1.8 s?
203Kcontext

Rankingin thinking mode?

Strong at

S
Tool calling#9 of 242Instructions#14 of 245
A
Reasoning#80 of 334Science#61 of 312Knowledge#65 of 318
B
Coding#49 of 132Agents & work tasks#62 of 175Long context#97 of 273

Alternatives?

ModelScoreCostPer taskWhat changes
CheaperMuse Spark 1.3max+98%-7%$1.60Only auto tool_choice is supported+5OpenGLM 5.3unstated+76%same$1.40 / $4.40Reasoning cannot be disabled+3CheaperMiMo v2.6 ProunstatedOpen+69%-75%$0.43 / $0.87Memory (4-bit): 452 GB → 615 GB+4CheaperGLM 5.3 FlashunstatedOpen+62%-89%$0.15 / $0.50Defaults to max · Self-hosted+4CheaperDeepSeek V4.1 Flash 1017unstatedOpen+62%-76%$0.30 / $1.20Memory (4-bit): 452 GB → 458 GB+5

Run it yourself

estimate
Fits on✗ 8 GB (phone, laptop)✗ 16 GB✗ 24 GB GPU✗ 48 GB GPU✗ 80 GB GPU
LicenceMITfree for commercial use

History

10 versions
253035404550Jul 26Release dateVals Legal Research Bench5.15.25.3
253035404550Jul 26Release dateVals Legal Research Bench5.15.25.3

Benchmarks

all categories
Chat & Knowledge
Humanity's Last Exam30.1#80
AA Omniscience Index0.9#65
CritPt (physics)4.6#81
AA Intelligence Index 4.3.226.1#75
AA Omniscience · Accuracy23.7#124
AA Omniscience · Hallucination Rate29.9#34
AA-LCR (long context) 1.173.7#97
GDP.pdf · AA all-pass8.4#66
SimpleQA-Verifiedlab38.1
Chinese-SimpleQAlab75
MMLU-Pro EMlab86
HMMT February 2026lab89.4
IMOAnswerBenchlab83.8
Apexlab11.5
Apex Shortlistlab72.4
GDPval-AA Elolab1,535
GPQA Diamond86.8#61
Coding
Terminal-Bench 4.02#49
SciCode44.8#75
Terminal-Bench Hard (version unstated)43.2#25
SWE-bench Pro (edition unstated)lab58.4
DeepSWE (edition unstated)lab18
Terminal Bench 2.0lab63.5
SWE Multilinguallab73.3
Terminal-Bench 2.161.8#50
Agents & Tool Use
AutomationBench-AA20.3#60
GDPval-AA (work tasks) 2.130.8#62
τ³-Banking · AA13.6#75
ITBench SRE40.3#17
AA-Briefcase 1.1 · Elo964.5#58
Toolathlonlab40.7
BrowseComplab79.3
HLE without toolslab34.7
HLE with toolslab50.4
MCPAtlas Publiclab71.8
τ²-bench · AA97.7#9
Writing & Language
IFBench (instructions)76.3#14

FAQ

How much does GLM 5.1 cost?›

$1.40 per 1M input tokens and $4.40 per 1M output tokens; cached input $0.26 per 1M; cheapest via lab-qwen at $0.82 / $3.30. A typical benchmark task costs about $1.02 at default effort.

Is GLM 5.1 free to use?›

There is no free API; the cheapest paid price is $0.82 per 1M input tokens. The weights are open, so you can run it yourself.

How good is GLM 5.1, and what is it best at?›

#80 of 334 in Chat & Knowledge, #49 of 132 in Coding, #60 of 133 in Agents & Tool Use, #29 of 195 in Writing & Language, #66 of 133 in Vision & Documents. Strongest at: Tool calling, Reasoning, Science, Knowledge, Instructions.

Is GLM 5.1 better than Inkling?›

On Vals Legal Research Bench, GLM 5.1 scores 27.9 vs 28.4 — about the same.

How can I use GLM 5.1?›

Through the API of Alibaba, AtlasCloud, Baidu, Chutes, DeepInfra and 6 more.

What is the context window of GLM 5.1?›

203K tokens, with up to 131K output tokens.

When was GLM 5.1 released? Is there a newer version?›

Released 2026-04-07. It has been superseded by GLM 5.2.

Is GLM 5.1 open source?›

Yes, the weights are open (MIT licence).

Can I run GLM 5.1 locally?›

The weights are open and can be self-hosted; hardware needs are not confirmed yet.

Can GLM 5.1 read images?›

No, it accepts text only.

What is GLM 5.1?›

GLM 5.1 is an open-weights language model from Z.ai, released Apr 2026, the successor to GLM 5.

What are the best alternatives to GLM 5.1?›

Muse Spark 1.3 (cheaper), GLM 5.3 (open weights).