Large Language ModelText Generation

DeepSeek R1-0528

Updated DeepSeek R1 with improved math reasoning

Released 2025-05-28

Context Window

128K

API Access

Available

Benchmark Performance

reasoning

reasoning

Massive Multitask Language Understanding

91.5%

math

math

MATH

98.0%

conversation

conversation

Chatbot Arena Elo

1416.0Elo