Megatron MarketsWholesale
Sign inContact Us
Models Model details

DeepSeek

DeepSeek: R1 0528

deepseek/deepseek-r1-0528

Enquire about this model Request a quote

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.

Input
$0.5

/ 1M tokens

Output
$2.15

/ 1M tokens

Context window
163.84K

tokens

Limits

Context

163,840

tokens

Max output

32,768

20% of the context window

Parameters

684,531,386,000

Model facts

Author

DeepSeek

Tokenizer

DeepSeek

Modality

Text -> Text

Instruction format

Deepseek-r1

License

MIT

Knowledge cutoff

2025-03-31

Created

May 28, 2025

Hugging Face

deepseek-ai/DeepSeek-R1-0528

Reasoning

Required

Publication status

Hugging Face: verifiedLicense: publishedParameter count: published

Modalities

Input · TextOutput · Text

Generation controls

Frequency penaltyInclude reasoningLogit biasLogprobsMax tokensMin pPresence penaltyReasoningRepetition penaltyResponse formatSeedStopStructured outputsTemperatureTool choiceToolsTop kTop logprobsTop p

Capability scores

13.1
Intelligence index

Better than 34% of models compared

Creative tasks

Category performance

Data visualization
Top 56%

1182 Elo

60.9% win

UI components
Top 75%

1102 Elo

54.9% win

3D
Top 67%

1112 Elo

53.3% win

Websites
Top 68%

1151 Elo

52.7% win

Code
Top 72%

1140 Elo

52.6% win

Games
Top 74%

1099 Elo

49.3% win

SVG
Top 86%

1038 Elo

48.7% win