Advertise here
#193

GLM 5.3 FlashX

Z.ai Release: 2026-09-21 Tested on: 2026-09-21 21:14 z-ai/glm-5.3-flashx::medium
320B total (18B active)MoEClosed
(max) (high) (medium) (low)

Summary

GLM 5.3 FlashX scores 6.4 on AI BENCHY and ranks #193. It has 10.0 reliability, a 68.2% pass rate, $0.056 total cost, and 4.32s average response time.

What makes GLM 5.3 FlashX unique: It stands out most in Coding, where it ranks #4, while General Intelligence is its weakest area at #14. It is notably fast compared with similar models.

Model facts

Researched on 2026-09-21

Reported
Parameters
320B total (18B active)
Architecture
MoE
Availability
Closed
License
-

OpenRouter reports 320B total and 18B active parameters for FlashX. This record describes the hosted FlashX checkpoint; the open Flash weights do not establish an exact FlashX weights release.

Score

6.4

Consistency

7.0

Total Output Tokens

19,433

Total Input Tokens

84,329

Input Price

$0.370 / 1M

Output Price

$1.250 / 1M

Tests Correct

Wrong Tests: 11

Attempt pass rate: 68.2%

Flaky tests

8

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

Response Time (avg)

4.32s

Response Time (max): 22.89s

Response Time (total): 95.15s

Hamster playing table tennis

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

#193 GLM 5.3 FlashX

medium
Cost
$0.002
Time
8.1s
Tokens
1,480 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

Category Breakdown

Category Score Consistency Tests Correct
Anti-AI Tricks 7.3 5.8
Coding 6.0 4.6
Combined 2.8 1.6
Data parsing and extraction 10.0 10.0
Domain specific 3.6 7.2
General Intelligence 3.4 2.2
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