DeepSeek-V4-ProvsDeepSeek-R1-0528

Across 5 shared benchmarks, DeepSeek-R1-0528 leads overall: DeepSeek-V4-Pro wins 1, DeepSeek-R1-0528 wins 4, with 0 ties and an average score difference of -4.14.

DeepSeek-AI
DeepSeek-V4-Pro

DeepSeek-AI · 2026-04-24 · Reasoning model

DeepSeek-AI
DeepSeek-R1-0528

DeepSeek-AI · 2025-05-28 · Reasoning model

DeepSeek-V4-Pro1 win(20%)(80%)4 winsDeepSeek-R1-0528

Benchmark scores

Grouped by capability, sorted by largest gap within each. 5 shared benchmarks.

General Knowledge

DeepSeek-R1-0528 3/3
BenchmarkDeepSeek-V4-ProDeepSeek-R1-0528Diff
HLE7.70141 / 157Normal (No Tools)17.70111 / 157-10
GPQA Diamond72.90102 / 178Normal (No Tools)8169 / 178-8.10
MMLU Pro82.9046 / 126Normal (No Tools)8525 / 126-2.10

Coding and Software Engineer

Even 2/2
BenchmarkDeepSeek-V4-ProDeepSeek-R1-0528Diff
LiveCodeBench56.8075 / 120Normal (No Tools)73.3045 / 120-16.50
SWE-bench Verified73.6041 / 108Normal (With Tools)57.6080 / 108+16

Specs

FieldDeepSeek-V4-ProDeepSeek-R1-0528
PublisherDeepSeek-AIDeepSeek-AI
Release date2026-04-242025-05-28
Model typeReasoning modelReasoning model
ArchitectureMoEMoE
Parameters1.6T671B
Context length1M64K
Max output375K64K

API pricing

Prices use DataLearner records when available; missing fields are not inferred.

ItemDeepSeek-V4-ProDeepSeek-R1-0528
Text input$0.435 / 1M tokensNot public
Text output$0.87 / 1M tokensNot public
Cache read$0.87 / 1M tokensNot public
Cache write$0.003625 / 1M tokensNot public

One or both models have incomplete public pricing.

Summary

  • DeepSeek-R1-0528leads in:General Knowledge (3/3)
  • Tied in:Coding and Software Engineer

On average across the 5 shared benchmarks, DeepSeek-R1-0528 scores 4.14 higher.

Largest single-benchmark gap: LiveCodeBench — DeepSeek-V4-Pro 56.80 vs DeepSeek-R1-0528 73.30 (-16.50).

Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.