DataLearner logo
GL

GLM-5.3-Flash

Multimodal modelCoding modelGLM FlashGLM-5

GLM-5.3-Flash

Also known as: Ox Alpha / ox-alpha / stealth/ox-alpha / GLM 5.3 Flash

Release date: 2026-08-26Updated: 2026-08-26 23:23:52.208242
Parameters
320B
Context length
1M
Chinese support
Supported
Reasoning ability

GLM-5.3-Flash is an AI model published by Zhipu AI, released on 2026-08-26, for Multimodal model, with 320B parameters, and 1M context length, with a 1773.00 score on GDPval-AA v2.

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

GLM-5.3-Flash

Model basics

Reasoning traces
Supported
Thinking modes
Thinking Level · Max (Default)Thinking Level · LowThinking Level · High
Context length
1M tokens
Max output length
128K tokens
Model type
Multimodal model
Modality (in / out)
Text, Image → Text
Release date
2026-08-26
Model file size
No data
MoE architecture
Yes
Total params / Active params
320B / 18B
Knowledge cutoff
No data
GLM-5.3-Flash

Open source & experience

Code license
Weights license
MIT License- Commercial use permitted
GLM-5.3-Flash

Official resources

DataLearnerAI blog
No blog post yet
GLM-5.3-Flash

API details

API speed
5/5
💡Default unit: $/1M tokens. If vendors use other units, follow their published pricing.
Standard
TypeConditionInputOutput
Text-$0.075/ 1M$0.250/ 1M
Cache PricingPrompt Cache
TypeTTLWriteRead
Text--$0.015/ 1M
GLM-5.3-Flash

Benchmark Results

GLM-5.3-Flash currently shows benchmark results led by HLE (15 / 183, score 55.30), AutomationBench (1 / 9, score 48.80), GDPval-AA v2 (2 / 15, score 1773). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.

Thinking
Tool usage

General Knowledge

1 evaluations
Benchmark / mode
Score
Rank/total
HLE
MaxTools
55.30
15 / 183

AI Agent - Tool Usage

3 evaluations
Benchmark / mode
Score
Rank/total
84.30
11 / 46
78.40
1 / 7
48.80
1 / 9

Coding and Software Engineer

2 evaluations
Benchmark / mode
Score
Rank/total
DeepSWE
MaxTools
63.40
11 / 30
56.30
4 / 10

Agent Level Benchmark

1 evaluations
Benchmark / mode
Score
Rank/total
26.30
8 / 13

Productivity Knowledge

2 evaluations
Benchmark / mode
Score
Rank/total
GDPval-AA v2
MaxTools
1773
2 / 15
62.40
2 / 2

Multimodal Understanding

4 evaluations
Benchmark / mode
Score
Rank/total
CharXiv RQ
MaxTools
89.40
4 / 18
MMVU
Max
80.50
2 / 2
Chartography
MaxTools
78
1 / 2
53.40
5 / 5

Compare with other models

GLM-5.3-Flash

Publisher

GLM-5.3-Flash

Model Overview

GLM-5.3-Flash is an AI model published by Zhipu AI, released on 2026-08-26, for Multimodal model, with 320B parameters, and 1M context length, with a 1773.00 score on GDPval-AA v2.

DataLearner on WeChat

Follow DataLearner on WeChat for AI model updates and research notes.

DataLearner WeChat QR code