Noometry

Model comparison

DeepSeek V4 Pro vs Qwen2.5-VL 72B Instruct

DeepSeek V4 Pro is the stronger model overall, scoring 54.2 to 29.4 on the Noometry Index.

Last verified . 1 shared benchmarks.

DeepSeek V4 Pro DeepSeek

54.2

Rank #28 Confirmed

Qwen2.5-VL 72B Instruct Alibaba (Qwen)

29.4

Rank #298 Confirmed

Summary

  • They share 1 benchmark with published results for both. DeepSeek V4 Pro scores higher in 2 categories and Qwen2.5-VL 72B Instruct in 0 categories; 2 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where DeepSeek V4 Pro leads 56.4 to 20.0.
  • The biggest single-benchmark swing is Kagi LLM Benchmark: 53.5% for DeepSeek V4 Pro and 36% for Qwen2.5-VL 72B Instruct.
  • DeepSeek V4 Pro is cheaper at $0.43 / $0.87 per million input/output tokens, against $2.80 / $8.40 for Qwen2.5-VL 72B Instruct.
  • DeepSeek V4 Pro accepts more context: 1M tokens versus 131K.

Side by side

DeepSeek V4 ProQwen2.5-VL 72B Instruct
ProviderDeepSeekAlibaba (Qwen)
Noometry Index54.229.4
Released2026-04-242024-09
WeightsOpenOpen
Context window1M131K
Max output393K8K
Input $ / M tokens$0.43$2.80
Output $ / M tokens$0.87$8.40
Results tracked486

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

Category by category

Coding Not comparable

DeepSeek V4 Pro: 52.2 (#31), Qwen2.5-VL 72B Instruct: —

BenchmarkDeepSeek V4 ProQwen2.5-VL 72B Instruct
SWE-bench Verified77.6%—
FrontierCode28.6%—
LMArena WebDev1582—
SciCode51%—
WeirdML66.2%—
LMArena Coding1470—
ALE-Bench1,403—

Agentic & Tool Use DeepSeek V4 Pro leads

DeepSeek V4 Pro: 32.7 (#54), Qwen2.5-VL 72B Instruct: 18.4 (#136)

BenchmarkDeepSeek V4 ProQwen2.5-VL 72B Instruct
APEX-Agents47.3%—
OSWorld—5%
Vending-Bench 23,285—

Reasoning DeepSeek V4 Pro leads

DeepSeek V4 Pro: 56.4 (#22), Qwen2.5-VL 72B Instruct: 20.0 (#228)

BenchmarkDeepSeek V4 ProQwen2.5-VL 72B Instruct
Kagi LLM Benchmark53.5%36%
ARC-AGI-261.3%—
NYT Connections (extended)91.3%—
ARC-AGI-190.5%—
CritPt18%—
Chess Puzzles47%—
LMArena Hard Prompts1461—
Mystery Game Puzzles43%—
DTBench93.9%—
LMCA45.5%—
Surface Evolver Bench40%—
Epoch Capabilities Index155.39—
ForecastBench56.1—

Math Not comparable

DeepSeek V4 Pro: 64.6 (#24), Qwen2.5-VL 72B Instruct: —

BenchmarkDeepSeek V4 ProQwen2.5-VL 72B Instruct
FrontierMath (Tiers 1-3)64.6%—
FrontierMath Tier 426.8%—
MathArena Final-Answer Competitions76.6%—
OTIS Mock AIME 2024-202598.6%—
ProofBench50%—
LMArena Math1455—

Knowledge Not comparable

DeepSeek V4 Pro: 59.5 (#28), Qwen2.5-VL 72B Instruct: —

BenchmarkDeepSeek V4 ProQwen2.5-VL 72B Instruct
GPQA Diamond91.7%—
SimpleQA Verified52.9%—
Vectara Hallucination Rate8.6%—
LMArena Expert1463—

Multimodal Not comparable

DeepSeek V4 Pro: —, Qwen2.5-VL 72B Instruct: 32.9 (#92)

BenchmarkDeepSeek V4 ProQwen2.5-VL 72B Instruct
LMArena Vision—1107
Video-MME—73.5%
GeoBench—62%
SpatialViz-Bench—33.3%

Multilingual Not comparable

DeepSeek V4 Pro: 54.5 (#44), Qwen2.5-VL 72B Instruct: —

BenchmarkDeepSeek V4 ProQwen2.5-VL 72B Instruct
LMArena Non-English1438—
LMArena Chinese1486—
LMArena French1473—
LMArena German1458—
LMArena Japanese1448—
LMArena Korean1452—
LMArena Russian1452—
LMArena Spanish1457—

Instruction Following Not comparable

DeepSeek V4 Pro: 76.3 (#44), Qwen2.5-VL 72B Instruct: —

BenchmarkDeepSeek V4 ProQwen2.5-VL 72B Instruct
LMArena Instruction Following1448—

Long Context Not comparable

DeepSeek V4 Pro: 45.0 (#49), Qwen2.5-VL 72B Instruct: —

BenchmarkDeepSeek V4 ProQwen2.5-VL 72B Instruct
CL-bench Life13.5%—
LMArena Longer Query1459—

Writing & Preference Not comparable

DeepSeek V4 Pro: 65.6 (#45), Qwen2.5-VL 72B Instruct: —

BenchmarkDeepSeek V4 ProQwen2.5-VL 72B Instruct
LMArena Text1451—
LMArena Creative Writing1446—
EQ-Bench Creative Writing1553—
EQ-Bench 41166—
LMArena Multi-Turn1467—

Frequently asked questions

Is DeepSeek V4 Pro better than Qwen2.5-VL 72B Instruct?

DeepSeek V4 Pro is the stronger model overall, scoring 54.2 to 29.4 on the Noometry Index.

Which is cheaper, DeepSeek V4 Pro or Qwen2.5-VL 72B Instruct?

DeepSeek V4 Pro is cheaper. It lists at $0.43 per million input tokens and $0.87 per million output tokens; Qwen2.5-VL 72B Instruct lists at $2.80 and $8.40.

Which has the bigger context window?

DeepSeek V4 Pro does, with 1M tokens against 131K.

How many benchmarks do DeepSeek V4 Pro and Qwen2.5-VL 72B Instruct share?

1 benchmark has published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Qwen2.5-VL 72B Instruct has 6.

Related comparisons

Go deeper