GLM 4.7

Dec 2025
GLM 4.7
Compare in
on
—per task?
$0.60 / $2.20per 1M in / out?cache $0.11
37 tok/sfirst token in 1.6 s?
203Kcontext

Rankingin thinking mode?

Strong at

S
Tool calling#12 of 242
A
Science#69 of 312Instructions#59 of 245
B
Reasoning#99 of 336Long context#119 of 314
C
Agents & work tasks#132 of 177Knowledge#165 of 320

Alternatives?

ModelScoreCostPer taskWhat changes
CheaperGPT-5.4 Nano (2026-03-17)high+6.3-54%$0.20 / $1.25BetterClaude Opus 4.8max+31.3+900%$4.08OpenGLM 5.1unstated+12.5+115%$1.40 / $4.40Memory (4-bit): 215 GB → 452 GBCheaperGPT-5 Mini (2025-08-07)high+6.3-31%$0.25 / $2CheaperGemini 2.5 Flashunstated+4.2-15%$0.30 / $2.50

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
0510Oct 25Jan 26Apr 26Release dateFrontierMath Tier 44.64.755.1
0510Oct 25Jan 26Apr 26Release dateFrontierMath Tier 44.64.755.1

Benchmarks

all categories
Chat & Knowledge
Humanity's Last Exam27.4#99
AA Omniscience Index-36.4#166
CritPt (physics)1.7#113
AA Intelligence Index 4.3.2≈22.2#104
AA Omniscience · Accuracy29.3#85
AA Omniscience · Hallucination Rate93#287
AA-LCR (long context) 1.171#119
GPQA Diamond85.9#69
AIME 2025 · AA archived95#4
Coding
Terminal-Bench Hard (version unstated)31.8#56
Terminal-Bench 2.145.3#75
LiveCodeBench · AA archived89.4#3
Agents & Tool Use
GDPval-AA (work tasks) 2.125.7#132
τ³-Banking · AA12.2#82
τ²-bench · AA95.9#12
Writing & Language
IFBench (instructions)67.9#59

FAQ

How much does GLM 4.7 cost?›

$0.60 per 1M input tokens and $2.20 per 1M output tokens; cached input $0.11 per 1M; cheapest via DeepInfra at $0.40 / $1.75.

Is GLM 4.7 free to use?›

There is no free API; the cheapest paid price is $0.40 per 1M input tokens. The weights are open, so you can run it yourself.

How good is GLM 4.7, and what is it best at?›

#99 of 336 in Chat & Knowledge, #58 of 97 in Coding, #132 of 177 in Agents & Tool Use, #62 of 197 in Writing & Language. Strongest at: Tool calling, Science, Instructions.

Is GLM 4.7 better than Qwen3.6 Flash?›

On FrontierMath Tier 4, GLM 4.7 scores 0 vs 0 — about the same.

How can I use GLM 4.7?›

Through the API of AWS Bedrock, DeepInfra, Mancer 2, Novita.

What is the context window of GLM 4.7?›

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

When was GLM 4.7 released? Is there a newer version?›

Released 2025-12-22. It has been superseded by GLM 5.

Is GLM 4.7 open source?›

Yes, the weights are open (MIT licence).

Can I run GLM 4.7 locally?›

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

Can GLM 4.7 read images?›

No, it accepts text only.

What is GLM 4.7?›

GLM 4.7 is an open-weights language model from Z.ai, released Dec 2025.

What are the best alternatives to GLM 4.7?›

GPT-5.4 Nano (2026-03-17) (cheaper), Claude Opus 4.8 (better), GLM 5.1 (open weights).