---
title: "GLM-5.3 Flash | Tarka Inference"
description: "Z.ai's efficient open model for high-volume chat and extraction."
url: "https://tarkahq.com/inference/glm-5-3-flash/"
---

Language model

# GLM-5.3 Flash

Z.ai's efficient open model for high-volume chat and extraction.

[Sign up](https://tarkahq.com/console?auth=signup) [Language model docs](https://tarkahq.com/docs/inference/llms/)

- **Status**: Available
- **Parameters**: 320B
- **Active**: 18B
- **Context**: 1M
- **Best for**: High-volume chat, extraction, and classification

## Model IDs

| Model ID | Use |
| --- | --- |
| `glm-5.3-flash` | High-volume chat, extraction, and classification |
| `glm-5.3-flash-eu` | European provider endpoint |

Use one of the model IDs above with Chat Completions or Responses at `https://tarka.rest/v1`.

## Pricing

`glm-5.3-flash`: $0.1485 input · $0.0297 cached input · $0.495 output per million tokens

`glm-5.3-flash-eu`: $0.1485 input · $0.0297 cached input · $0.495 output per million tokens

 USD · catalog 2026-11-04.1. Release verified 2026-11-04. EUR uses the catalog conversion. [Live catalog](https://tarka.rest/control/v1/pricing).

## Overview

GLM-5.3 Flash is Z.ai’s open-weight mixture-of-experts model, released under the MIT license. It has 320 billion parameters, 18 billion of them active for each token, and a context of about a million tokens. It reasons before it answers by default.

## When to pick it

- High-volume chat, summarization, and extraction
- Classification and routing steps inside an agent
- Long documents and transcripts that fit in its context

## When to look elsewhere

- The hardest coding and agent work: try Kimi K3
- Fast general-purpose chat with tool use: compare DeepSeek V4.1 Flash

## In a tool loop

When a response carries `reasoning_content`, send it back with that assistant message in the next request, so the model keeps its earlier reasoning.