Model comparison

GLM-5.3-Flash vs Qwen3-4B

GLM-5.3-Flash is the stronger model overall, scoring 51.8 to 31.9 on the Noometry Index.

Last verified October 10, 2026. 3 shared benchmarks.

GLM-5.3-Flash Z.ai (Zhipu)

51.8

Rank #41 Confirmed

Qwen3-4B

Alibaba (Qwen)

31.9

Rank #264 Confirmed

Summary

They share 3 benchmarks with published results for both. GLM-5.3-Flash scores higher in 4 categories and Qwen3-4B in 0 categories; 4 gaps are clear of the uncertainty.
The widest gap is in reasoning, where GLM-5.3-Flash leads 48.0 to 19.2.
The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 93.9% for GLM-5.3-Flash and 52.2% for Qwen3-4B.

Side by side

GLM-5.3-Flash and Qwen3-4B specifications
	GLM-5.3-Flash	Qwen3-4B
Provider	Z.ai (Zhipu)	Alibaba (Qwen)
Noometry Index	51.8	31.9
Released	2026-08-20	2025-04-29
Weights	Open	Open
Context window	1M	—
Max output	131K	—
Input $ / M tokens	$0.15	—
Output $ / M tokens	$0.50	—
Results tracked	40	6

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Not comparable

GLM-5.3-Flash: 53.1 (#31), Qwen3-4B: —

Coding benchmarks
Benchmark	GLM-5.3-Flash	Qwen3-4B
DeepSWE	63.4%	—
FrontierCode	31.8%	—
CursorBench	36.8%	—
LMArena WebDev	1609	—
FrontierSWE	18.1%	—
SciCode	51.6%	—
LMArena Coding	1508	—
ALE-Bench	303.55	—

Agentic & Tool Use GLM-5.3-Flash leads

GLM-5.3-Flash: 34.2 (#47), Qwen3-4B: 27.6 (#100)

Agentic & Tool Use benchmarks
Benchmark	GLM-5.3-Flash	Qwen3-4B
APEX-Agents	52.8%	—
Berkeley Function Calling Leaderboard	—	35.7%
GDP.pdf	14%	—

Reasoning GLM-5.3-Flash leads

GLM-5.3-Flash: 48.0 (#42), Qwen3-4B: 19.2 (#268)

Reasoning benchmarks
Benchmark	GLM-5.3-Flash	Qwen3-4B
Chess Puzzles	14%	4%
ARC-AGI-2	65.8%	—
ARC-AGI-1	91%	—
CritPt	15.4%	—
LMArena Hard Prompts	1491	—
Mystery Game Puzzles	8%	—
Surface Evolver Bench	52.5%	—
Bench to the Future 3	0.15	—
Epoch Capabilities Index	151.88	—

Math GLM-5.3-Flash leads

GLM-5.3-Flash: 53.3 (#47), Qwen3-4B: 29.7 (#240)

Math benchmarks
Benchmark	GLM-5.3-Flash	Qwen3-4B
OTIS Mock AIME 2024-2025	93.9%	52.2%
FrontierMath (Tiers 1-3)	55.8%	—
FrontierMath Tier 4	17.1%	—
MathArena Final-Answer Competitions	—	38.5%
ProofBench	21%	—
LMArena Math	1500	—

Knowledge GLM-5.3-Flash leads

GLM-5.3-Flash: 58.4 (#36), Qwen3-4B: 33.0 (#208)

Knowledge benchmarks
Benchmark	GLM-5.3-Flash	Qwen3-4B
GPQA Diamond	90.2%	52.3%
Vectara Hallucination Rate	—	5.7%
LMArena Expert	1513	—

Multimodal Not comparable

GLM-5.3-Flash: 42.8 (#27), Qwen3-4B: —

Multimodal benchmarks
Benchmark	GLM-5.3-Flash	Qwen3-4B
LMArena Vision	1296	—

Multilingual Not comparable

GLM-5.3-Flash: 56.0 (#25), Qwen3-4B: —

Multilingual benchmarks
Benchmark	GLM-5.3-Flash	Qwen3-4B
LMArena Non-English	1462	—
LMArena Chinese	1527	—
LMArena French	1496	—
LMArena German	1470	—
LMArena Japanese	1429	—
LMArena Korean	1446	—
LMArena Russian	1469	—
LMArena Spanish	1471	—

Instruction Following Not comparable

GLM-5.3-Flash: 77.5 (#20), Qwen3-4B: —

Instruction Following benchmarks
Benchmark	GLM-5.3-Flash	Qwen3-4B
LMArena Instruction Following	1478	—

Long Context Not comparable

GLM-5.3-Flash: 45.4 (#39), Qwen3-4B: —

Long Context benchmarks
Benchmark	GLM-5.3-Flash	Qwen3-4B
LMArena Longer Query	1482	—

Writing & Preference Not comparable

GLM-5.3-Flash: 65.3 (#50), Qwen3-4B: —

Writing & Preference benchmarks
Benchmark	GLM-5.3-Flash	Qwen3-4B
LMArena Text	1471	—
LMArena Creative Writing	1442	—
LMArena Multi-Turn	1467	—

Frequently asked questions

Is GLM-5.3-Flash better than Qwen3-4B?