DataLearner logo
DE

DeepSeek-R1

Reasoning modelDeepSeek R1DeepSeek R1

DeepSeek-R1

Release date: 2025-01-20Updated: 2025-03-21Views: 1,866
Live demoGitHubHugging FaceCompare
Parameters
671B
Context length
128K
Multilingual
Supported
Reasoning ability
No data

DeepSeek-R1 is an AI model published by DeepSeek-AI, released on 2025-01-20, for Reasoning model, with 671B parameters, and 128K context length, requiring about 134GB storage, with a 1500.00 score on Creative Writing.

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

DeepSeek-R1

Model basics

Reasoning traces
Supported
Thinking modes
Thinking modes not supported
Context length
128K tokens
Max output length
No data
Model type
Reasoning model
Modality (in / out)
No data
Release date
2025-01-20
Model file size
134GB
MoE architecture
No
Total params / Active params
671B / Not applicable
Knowledge cutoff
No data
DeepSeek-R1

Open source & experience

Code license
Weights license
MIT License- Commercial use permitted
GitHub repo
N/A
Live demo
N/A
DeepSeek-R1

Official resources

Paper
DataLearnerAI blog
DeepSeek-R1

API details

API speed
No data
💡Default unit: $/1M tokens. If vendors use other units, follow their published pricing.
Standard
TypeConditionInputOutput
Text-$0.550/ 1M tokens$2.19/ 1M tokens
Cache PricingPrompt Cache
TypeTTLWriteRead
Text-$0.550/ 1M tokens$0.140/ 1M tokens
Text-$0.550/ 1M tokens
Cache = write
Text-$0.140/ 1M tokens
Cache = hit

“—” means the modality is not billed in that direction, or the vendor has not published a price for it.

DeepSeek-R1

Benchmark Results

DeepSeek-R1 currently shows benchmark results led by MMLU (8 / 124, score 90.80), MMLU Pro (40 / 176, score 84), MATH-500 (13 / 45, score 97.30). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.

Thinking
Tool usage

General Knowledge

5 evaluations
Benchmark / mode
Score
Rank/total
MMLU
Standard Mode
90.80
8 / 124
MMLU Pro
Standard Mode
84
40 / 176
ARC-AGI-1
Standard Mode
15.80
135 / 147
HLE
Thinking Mode
8.50
412 / 563
CritPt
Thinking Mode
0.60
168 / 200

General Evaluation

2 evaluations
Benchmark / mode
Score
Rank/total
GPQA Diamond
Standard Mode
71.50
306 / 462
GPQA Diamond
Thinking Mode
70.80
317 / 462

Common Sense

1 evaluations
Benchmark / mode
Score
Rank/total
SimpleQA
Standard Mode
30.10
24 / 47

Coding and Software Engineer

5 evaluations
Benchmark / mode
Score
Rank/total
LiveCodeBench
Standard Mode
65.90
107 / 250
LiveCodeBench
Thinking Mode
61.70
126 / 250
SWE-bench Verified
Standard Mode
49.20
98 / 116
SciCode
Thinking Mode
38.30
110 / 130
WeirdML v2
Standard ModeTools
36.49
49 / 52

Math and Reasoning

5 evaluations
Benchmark / mode
Score
Rank/total
MATH-500
Standard Mode
97.30
13 / 45
AIME 2024
Standard Mode
79.80
28 / 62
AIME2025
Standard Mode
70
120 / 215
AIME2025
Thinking Mode
68
124 / 215
3.80
22 / 24

Writing and Creative Capabilities

1 evaluations
Benchmark / mode
Score
Rank/total
Creative Writing
Standard Mode
1500
51 / 106

Common Sense Reasoning

1 evaluations
Benchmark / mode
Score
Rank/total
SimpleBench
Standard Mode
30.90
75 / 93

Agent Level Benchmark

6 evaluations
Benchmark / mode
Score
Rank/total
Aider-Polyglot
Thinking Mode
56.90
25 / 59
BALROG
Standard ModeTools
34.90
8 / 12
METR Time Horizons v1.1
Standard ModeTools
26.93
21 / 22
τ²-Bench - Telecom
Thinking ModeTools
11.40
262 / 264
τ³-Banking
Thinking ModeTools
6.40
144 / 164
Terminal Bench Hard
Thinking ModeTools
6.10
204 / 244

Instruction Following

1 evaluations
Benchmark / mode
Score
Rank/total
IF Bench
Thinking Mode
39
223 / 282

AI Agent - Tool Usage

1 evaluations
Benchmark / mode
Score
Rank/total
Terminal-Bench 2.1
Thinking ModeTools
19.10
163 / 192

Long Context

1 evaluations
Benchmark / mode
Score
Rank/total
Fiction.liveBench
Standard Mode
69.40
11 / 16

Multimodal Understanding

1 evaluations
Benchmark / mode
Score
Rank/total
GDP.pdf
Thinking Mode
2.40
105 / 118
DeepSeek-R1

Publisher

DeepSeek-R1

Model Overview

DeepSeek-R1 is an AI model published by DeepSeek-AI, released on 2025-01-20, for Reasoning model, with 671B parameters, and 128K context length, requiring about 134GB storage, with a 1500.00 score on Creative Writing.

DataLearner on WeChat

Follow DataLearner on WeChat for AI model updates and research notes.

DataLearner WeChat QR code