AI BENCHY
Advertise here
#102

GLM 5.3 Flash

Z.ai Release: 2026-08-26 Tested on: 2026-08-26 17:11 z-ai/glm-5.3-flash::high
320B total (18B active)MoEOpen source
(max) (high) (low)

Summary

GLM 5.3 Flash scores 7.3 on AI BENCHY and ranks #102. It has 10.0 reliability, a 66.7% pass rate, $0.029 total cost, and 25.36s average response time.

What makes GLM 5.3 Flash unique: It stands out most in Instructions following, where it ranks #1, while Data parsing and extraction is its weakest area at #15. Its total benchmark cost is unusually low for its score range.

Model facts

Researched on 2026-08-26

Reported
Parameters
320B total (18B active)
Architecture
MoE
Availability
Open source
License
MIT

Z.ai reports a hybrid sparse and linear attention architecture for this native multimodal checkpoint.

Parameter source 1 Parameter source 2 Availability source

Score

7.3

Consistency

8.9

Reliability

10.0

Total Cost (Current Price)

$0.029

Total Output Tokens

78,318

Total Input Tokens

113,003

Input Price

$0.075 / 1M

Output Price

$0.250 / 1M

Tests Correct

Wrong Tests: 9

Attempt pass rate: 66.7%

Flaky tests

3

Flaky tests had mixed outcomes across runs (at least one pass and one fail).

Response Time (avg)

25.36s

Response Time (max): 163.41s

Response Time (total): 558.01s

Wrong answer: 4 Extra formatting: 2 Did not follow instructions: 1 Invalid tool call: 1 No answer: 1

Hamster playing table tennis

Prompt: Create a detailed SVG illustration of a hamster playing table tennis.

#102 GLM 5.3 Flash

high
Cost
$0.003
Time
186.7s
Tokens
10,486 tok

Charts

Choose the first model, then click a second model to open a side-by-side page.

Total Output Tokens

Score vs Total Output Tokens

Quick Compare

GLM 5.3 FlashhighvsKAT-Coder-Pro V2.5lowGLM 5.3 FlashhighvsGemini 3.1 Flash LitemediumGLM 5.3 FlashhighvsGLM 5.3highGLM 5.3 FlashhighvsQwen3.7 PlusnoneGLM 5.3 FlashhighvsClaude Sonnet 4.6noneGLM 5.3 FlashhighvsDeepSeek V4 Flash 0731mediumGLM 5.3 FlashhighvsGemini 3 Flash PreviewlowGLM 5.3 FlashhighvsClaude Opus 4.8noneGLM 5.3 FlashhighvsGrok 4.6lowGLM 5.3 FlashhighvsKimi K2.6medium

Category Breakdown

Category Score Consistency Tests Correct
Anti-AI Tricks 8.3 10.0
Coding 8.2 7.2
Combined 6.4 5.8
Data parsing and extraction 7.3 5.8
Domain specific 3.0 10.0
General Intelligence 5.0 10.0
Instructions following 10.0 10.0
Puzzle Solving 10.0 10.0
Tool Calling 10.0 10.0
Trivia 3.0 10.0

Compared models