Model specification · DeepSeek

R1

deepseek/deepseek-r1

DeepSeek R1 is here: 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....

SpecificationsReported values
Context window
64K
Maximum output
16K
Architecture
text->text
Tokenizer
DeepSeek
Knowledge cutoff
2024-07-31
Moderated
No
Input modalities
text
Output modalities
text
Full pricingUSD where reported
PriceStandard
Input$0.700 / M
Output$2.50 / M
Benchmark indexesReported indexes
Intelligence
11.4
Coding
24.6
Agentic
1.1
Design Arena Elo
—
Example workload costsStandard pricing only
WorkloadInput / outputCache shareEstimated cost
Quick check10K / 1K × 10%$0.00950
Long analysis100K / 4K × 150%$0.080
Workday20K / 2K × 1,00030%$19.00
More from DeepSeekAll DeepSeek models
ModelContextInput / MOutput / MIntel.
DeepSeek V4.1 Flash1.0M$0.150$0.60039.5
DeepSeek V4 Pro 08131.0M$0.578$1.7336.3
DeepSeek V4 Flash 07311.3M$0.040$0.08034.5
DeepSeek V4 Pro 04231.0M$0.442$0.88530.9
DeepSeek V4 Flash 04231.0M$0.044$0.08824.8
DeepSeek V3.1 Terminus164K$0.270$1.0015.4
DeepSeek V3 0324164K$0.250$1.009.7