# GLM 5.3 Flash

Fast, cost-optimized tier of the GLM 5.3 family. Native multimodal (text, image, video input) with a hybrid sparse/linear attention architecture for efficient coding and long-horizon agent tasks.

## Overview

- **Provider:** ZAI
- **Model ID:** `glm-5.3-flash`
- **Type:** Chat
- **Canonical URL:** https://modelstack.cc/models/chat/glm-5.3-flash
- **Version:** 5.3 Flash
- **Best for:** High-volume workloads, fast responses, cost-sensitive coding and agentic tasks

## Pricing

| Usage | Price |
| --- | ---: |
| Input | $0.15 / 1M tokens |
| Output | $0.5 / 1M tokens |
| Cache read | $0.03 / 1M tokens |
| Cache write (5 minutes) | $0.15 / 1M tokens |

## Specifications

- **Context window:** 1.25M
- **Maximum output:** 131K
- **Released:** 2026-08-26

## Capabilities

- chat
- tool_use
- reasoning
- coding
- analysis

## API usage

Use this model ID in a ModelStack API request:

```json
{
  "model": "glm-5.3-flash"
}
```

For the rendered page, visit https://modelstack.cc/models/chat/glm-5.3-flash.