Model comparison

Qwen2.5-VL 72B Instruct vs Step 3

Step 3 is the stronger model overall, scoring 39.9 to 29.4 on the Noometry Index.

Last verified October 10, 2026. 2 shared benchmarks.

Qwen2.5-VL 72B Instruct

Alibaba (Qwen)

29.4

Rank #298 Confirmed

Step 3

StepFun

39.9

Rank #145 Confirmed

Summary

They share 2 benchmarks with published results for both. Qwen2.5-VL 72B Instruct scores higher in 0 categories and Step 3 in 2 categories; 2 gaps are clear of the uncertainty.
The widest gap is in reasoning, where Step 3 leads 27.6 to 20.0.
The biggest single-benchmark swing is Kagi LLM Benchmark: 36% for Qwen2.5-VL 72B Instruct and 62.3% for Step 3.

Side by side

Qwen2.5-VL 72B Instruct and Step 3 specifications
	Qwen2.5-VL 72B Instruct	Step 3
Provider	Alibaba (Qwen)	StepFun
Noometry Index	29.4	39.9
Released	2024-09	—
Weights	Open	Open
Context window	131K	—
Max output	8K	—
Input $ / M tokens	$2.80	—
Output $ / M tokens	$8.40	—
Results tracked	6	17

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

Category by category

Coding Not comparable

Qwen2.5-VL 72B Instruct: —, Step 3: 39.5 (#144)

Coding benchmarks
Benchmark	Qwen2.5-VL 72B Instruct	Step 3
LMArena Coding	—	1367

Agentic & Tool Use Not comparable

Qwen2.5-VL 72B Instruct: 18.4 (#136), Step 3: —

Agentic & Tool Use benchmarks
Benchmark	Qwen2.5-VL 72B Instruct	Step 3
OSWorld	5%	—

Reasoning Step 3 leads

Qwen2.5-VL 72B Instruct: 20.0 (#228), Step 3: 27.6 (#102)

Reasoning benchmarks
Benchmark	Qwen2.5-VL 72B Instruct	Step 3
Kagi LLM Benchmark	36%	62.3%
LMArena Hard Prompts	—	1355

Math Not comparable

Qwen2.5-VL 72B Instruct: —, Step 3: 34.9 (#148)

Math benchmarks
Benchmark	Qwen2.5-VL 72B Instruct	Step 3
LMArena Math	—	1366

Knowledge Not comparable

Qwen2.5-VL 72B Instruct: —, Step 3: 36.2 (#161)

Knowledge benchmarks
Benchmark	Qwen2.5-VL 72B Instruct	Step 3
LMArena Expert	—	1333

Multimodal Step 3 leads

Qwen2.5-VL 72B Instruct: 32.9 (#92), Step 3: 34.9 (#81)

Multimodal benchmarks
Benchmark	Qwen2.5-VL 72B Instruct	Step 3
LMArena Vision	1107	1176
Video-MME	73.5%	—
GeoBench	62%	—
SpatialViz-Bench	33.3%	—

Multilingual Not comparable

Qwen2.5-VL 72B Instruct: —, Step 3: 46.5 (#154)

Multilingual benchmarks
Benchmark	Qwen2.5-VL 72B Instruct	Step 3
LMArena Non-English	—	1327
LMArena Chinese	—	1396
LMArena German	—	1371
LMArena Korean	—	1271
LMArena Russian	—	1331
LMArena Spanish	—	1372

Instruction Following Not comparable

Qwen2.5-VL 72B Instruct: —, Step 3: 70.6 (#159)

Instruction Following benchmarks
Benchmark	Qwen2.5-VL 72B Instruct	Step 3
LMArena Instruction Following	—	1332

Long Context Not comparable

Qwen2.5-VL 72B Instruct: —, Step 3: 40.3 (#152)

Long Context benchmarks
Benchmark	Qwen2.5-VL 72B Instruct	Step 3
LMArena Longer Query	—	1326

Writing & Preference Not comparable

Qwen2.5-VL 72B Instruct: —, Step 3: 54.5 (#146)

Writing & Preference benchmarks
Benchmark	Qwen2.5-VL 72B Instruct	Step 3
LMArena Text	—	1351
LMArena Creative Writing	—	1322
LMArena Multi-Turn	—	1340

Frequently asked questions

Is Qwen2.5-VL 72B Instruct better than Step 3?