DeepSeek

DeepSeek R1 0528 (May '25)

Unknown Size

By DeepSeek • Released 2025-05-28

Capability Radar

Avg Score
52

Across all benchmarks

Participated
13
Benchmarks

Benchmark Performance

Benchmark Category Score
MATH-500
Reasoning
98.3
MMLU-Pro
Knowledge
84.9
GPQA Diamond
Knowledge
81.3
LiveCodeBench
Coding
77
AIME 2025
Reasoning
76
LCR
Long-Context Reasoning
54.7
SciCode
Reasoning Knowledge
40.3
IFBench
Agent
39.6
𝜏²-Bench Telecom
Reasoning Knowledge
36.5
Artificial Analysis Intelligence Index
Knowledge
27
Artificial Analysis Coding Index
Coding
24
Terminal-Bench Hard
Agent Coding
15.9
HLE
Knowledge Multi-Modal
14.9