Back to Models
deepseek logo
deepseek/deepseek-r1-0528
Not Available

R1 0528

May 28th update to the [original DeepSeek R1](/deepseek/deepseek-r1) Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active in an inference pass. Fully open-source model.

5/28/2025
163,840 tokens
#79 Text (Japanese)

Specifications

Modalities

Input
text
Output
text

Supported Parameters

frequency_penalty
include_reasoning
logit_bias
logprobs
max_tokens
min_p
presence_penalty
reasoning
repetition_penalty
response_format
seed
stop
structured_outputs
temperature
tool_choice
tools
top_k
top_logprobs
top_p

Max Output Tokens

65,536

Reasoning Configuration

Default
Thinking on
Turning thinking off
Not possible (always thinks)

How to control thinking, and what it costs

Frequently asked questions

Is R1 0528 available on FastMetal?
Not at the moment. DeepSeek V4.1 Flash, from the same lab, is available on the FastMetal API today.
What is the context window of R1 0528?
163,840 tokens, shared between the prompt and the response.
How does R1 0528 rank?
#79 on the public arena's Japanese board (ELO 1,392). Ranks move as the leaderboard is updated.

Leaderboard

Text
OverallELO: 1,421
#115
JapaneseELO: 1,392
#79
ChineseELO: 1,429
#146
KoreanELO: 1,352
#117
EnglishELO: 1,433
#112
frenchELO: 1,417
#134
germanELO: 1,392
#120
spanishELO: 1,400
#130
russianELO: 1,412
#110
CodingELO: 1,464
#119
MathELO: 1,393
#155
Creative WritingELO: 1,390
#115
Instruction FollowingELO: 1,391
#144
Hard PromptsELO: 1,432
#126
Multi-TurnELO: 1,405
#141