GLM-5.3

GLM-5.3

The most capable open-weights model for coding, from Z.ai.

Z.ai's flagship open-weight reasoning model for coding agents and long-horizon tasks, with a 1M-token context window.

GLM-5.3 model card on Hugging Face

GLM-5.3 model card on Hugging Face

Overview

What is GLM-5.3?

The most capable open-weights model for coding, from Z.ai.

GLM-5.3 is the flagship model of Z.ai (formerly Zhipu AI), released in August 2026. It is a 753-billion-parameter mixture-of-experts model with 40 billion active parameters per token, built for coding agents and long-horizon engineering work. The model accepts text input only and offers a 1,048,576-token context window.

Its weights are publicly available on Hugging Face under Z.ai's custom GLM-5.3 License. On Artificial Analysis' independent Intelligence Index v5.0.1, GLM-5.3 scores 32 at maximum reasoning effort, with 41.9% on Terminal-Bench 4.0. Z.ai's API lists $1.40 per million input tokens and $4.40 per million output tokens.

Platforms and languages

GLM-5.3 Availability

Platforms

WebApi

Languages

EnglishChinese

Capabilities

GLM-5.3 Key Features

753B MoE reasoning model

753 billion total parameters with 40 billion active per token, built for extended reasoning and agent work.

1M-token context window

Handles very long codebases, documents and agent trajectories in a single context.

Open weights

Weights available on Hugging Face under the custom GLM-5.3 License for self-hosting.

Coding-agent performance

Scores 41.9% on Terminal-Bench 4.0 in Artificial Analysis independent testing (Index v5.0.1).

Available from 24 API providers

Served through Z.ai and other hosts, with hosted API pricing from $1.40 per million input tokens.

Plans and access

GLM-5.3 Pricing

Paid

Common questions

GLM-5.3 FAQs

How much does the GLM-5.3 API cost?

Z.ai lists $1.40 per million input tokens, $0.26 cached, and $4.40 per million output tokens.

Is GLM-5.3 open source?

Weights are public on Hugging Face, but under Z.ai's custom GLM-5.3 License, not a standard open-source licence.

What is GLM-5.3 best for?

Coding agents and long-horizon engineering tasks, where its Terminal-Bench results lead the GLM family.

Does GLM-5.3 accept images?

No. GLM-5.3 is text-only; the multimodal option in the family is GLM-5.3-Flash.

Reviews

See what the community thinks and share your experience.

—

Based on 0 ratings

Rating distribution

0
0
0
0
0

Leave a review

Sign in to rate

Community reviews

No reviews

No written reviews yet.

Explore by category

Llms Foundation Models

View all Llms Foundation Models websites
Qwen4-Max preview
Qwen4-Max

Alibaba's flagship 2.4T-parameter model with a 1M-token context and native vision, priced from $1.50 per million input tokens.

Llms Foundation Models
Sarvam Prism preview
Sarvam Prism

Sarvam AI's open-weight System One decision model that scores fixed options in a single pass and returns calibrated probabilities for enterprise workflows.

Llms Foundation Models

Explore by category

AI Coding Assistants

View all AI Coding Assistants websites
UI Verify preview
UI Verify

Visual regression testing for Playwright, Storybook, and Vitest that lets coding agents fix UI changes before merge.

AI Coding Assistants
infaiw preview
infaiw

A local-first open-source harness that runs coding agents in the background and only finishes work that passes your checks.

AI Coding Assistants