GLM-4.5

Name: GLM-4.5-MoE-355B-A32B-0715
Author: 智谱AI

Reasoning modelGLMGLM-4.5

GLM-4.5-MoE-355B-A32B-0715

Release date: 2025-07-28Updated: 2026-06-15 07:18:18.7871,630

Live demo GitHub Hugging Face Compare

Parameters

355B

Context length

128K

Chinese support

Supported

Reasoning ability

GLM-4.5-MoE-355B-A32B-0715 is an AI model published by 智谱AI, released on 2025-07-28, for Reasoning model, with 355B parameters, and 128K context length, requiring about 710 GB storage, under the Apache 2.0 license, with a 98.20 score on MATH-500.

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

GLM-4.5

Model basics

Reasoning traces

Supported

Thinking modes

Thinking modes not supported

Context length

128K tokens

Max output length

95K tokens

Model type

Reasoning model

Modality (in / out)

Text → Text

Release date

2025-07-28

Model file size

710 GB

MoE architecture

Yes

Total params / Active params

355B / 32B

Knowledge cutoff

No data

GLM-4.5

Open source & experience

Code license

Apache 2.0

Weights license

Apache 2.0- 免费商用授权

GitHub repo

https://github.com/THUDM/GLM-4

Hugging Face

https://huggingface.co/zai-org/GLM-4.5

Live demo

https://chat.z.ai/

GLM-4.5

Official resources

Paper

GLM-4.5: Reasoning, Coding, and Agentic Abililties

DataLearnerAI blog

GLM-4.5

API details

API speed

3/5

No public API pricing yet.

GLM-4.5

Benchmark Results

GLM-4.5 currently shows benchmark results led by MATH-500 (3 / 44, score 98.20), AIME 2024 (14 / 62, score 91), MMLU Pro (32 / 126, score 84.60). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.