DataLearner logoDataLearnerAI
Latest AI Insights
Model Evaluations
Model Directory
Model Comparison
Resource Center
Tools
LanguageEnglish

加载中...

DataLearner logoDataLearner AI

A knowledge platform focused on LLM benchmarking, datasets, and practical instruction with continuously updated capability maps.

Products

  • Leaderboards
  • Model comparison
  • Datasets

Resources

  • Tutorials
  • Editorial
  • Tool directory

Company

  • About
  • Privacy policy
  • Data methodology
  • Contact

© 2026 DataLearner AI. DataLearner curates industry data and case studies so researchers, enterprises, and developers can rely on trustworthy intelligence.

Privacy policyTerms of service
Loading comparison...
Table of Contents
目录
HomeModel CompareDeepSeek-V3-0324 vs GPT-4.5 vs DeepSeek-V3 vs DeepSeek-R1 评测对比

大模型评测对比结果

See key specs and per-benchmark scores for each model/mode. Scroll horizontally for all columns. 当前对比 4 个模型的评测数据与核心参数。

DeepSeek-V3-0324GPT-4.5DeepSeek-V3DeepSeek-R1
规格对比
DeepSeek-AI

DeepSeek-V3-0324

DE

DeepSeek-V3-0324

Release2025-03-24
Context length128K
Parameters6710
常规模式（Non-Thinking Mode）
Model profilePlayground
OpenAI

GPT-4.5

GP

GPT-4.5

Release2025-02-28
Context length128K
Parameters0
聊天大模型
Model profile
DeepSeek-AI

DeepSeek-V3

DE

DeepSeek-V3

Release2024-12-26
Context length128K
Parameters6810
聊天大模型
Model profile
DeepSeek-AI

DeepSeek-R1

DE

DeepSeek-R1

Release2025-01-20
Context length128K
Parameters6710
推理大模型
Model profile

Performance benchmarks

Compare benchmark results across thinking modes and tool usage.

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

Thinking
Tool usage
Internet

Best Overall

—

Best Single

—

Thinking Mode (Default)

DeepSeek-V3-0324 · 1 Modality support

Benchmark scores

Higher is usually better; “—” means no score.

Filter: Default Mode0 All Modes · 0 Benchmark
No data

Benchmark score table

Complete scores for each model/mode across selected benchmarks.

Feature compare

Detailed feature breakdown

Licensing, MoE architecture, and multi-modality support.

Features & specs
DE
DeepSeek-V3-0324DeepSeek-AI
GP
GPT-4.5OpenAI
DE
DeepSeek-V3DeepSeek-AI
DE
DeepSeek-R1DeepSeek-AI

Model snapshots

Organization
DeepSeek-AIOpenAIDeepSeek-AIDeepSeek-AI
模型全名
DeepSeek-V3-0324GPT-4.5DeepSeek-V3DeepSeek-R1
模型简介
Not providedNot providedNot providedNot provided
模型类型
聊天大模型聊天大模型聊天大模型推理大模型
模型代号
DeepSeek-V3-0324gpt-4_5DeepSeek-V3DeepSeek-R1
Release
2025-03-242025-02-282024-12-262025-01-20
MoE
YesNoNoNo

规格与性能

Context length
128K128K128K128K
Parameters
6710—68106710
激活参数量
370Not providedNot providedNot provided
模型规模
100b100b100b100b
模型大小
1442GBNot provided687.9 GB134GB
推理速度
Not providedNot providedNot provided
推理等级
Not providedNot providedNot provided
最大输出
Not providedNot providedNot providedNot provided
Supported modes
常规模式（Non-Thinking Mode）
No mode dataNo mode dataNo mode data

开源与许可

Code Open Source
Closed SourceNot providedClosed SourceClosed Source
Weights Open Source
Closed SourceNot providedNot providedClosed Source
Commercial use
免费商用授权不开源免费商用授权免费商用授权

Modality support

Text Input/Output
/
Not providedNot providedNot provided
Image Input/Output
/
Not providedNot providedNot provided
Audio Input/Output
/
Not providedNot providedNot provided
Video Input/Output
/
Not providedNot providedNot provided
Embedding Input/Output
/
Not providedNot providedNot provided

API 接口详情

Text 价格
Input: 0.27 美元/100万 tokensOutput: 1.1 美元/100万 tokens
Not providedNot providedNot provided
Image API pricing
Not providedNot providedNot providedNot provided
Audio API pricing
Not providedNot providedNot providedNot provided
Video API pricing
Not providedNot providedNot providedNot provided
Embedding API pricing
Not providedNot providedNot providedNot provided

Resources

GitHub
RepoNot providedRepoNot provided
Hugging Face
Model PageNot providedModel PageModel Page
Official Page
Not providedNot providedNot providedNot provided
Guides
Not providedNot providedNot providedNot provided
Papers
Introducing DeepSeek-V3DeepSeek_R1.pdf
DataLearnerAI
DeepSeekV3-0324发布：DeepSeek V3基础上大幅升级推理能力和前端网页的美观度，多项评测结果超过GPT-4.5Not provided开源大模型的新里程碑：DeepSeek AI开源6510亿参数的DeepSeek V3模型，评测结果显著好于4050亿参数的Llama3.1 405B，比肩Sonnet 3.5的开源模型什么是推理大模型？DeepSeek R1推理大模型与DeepSeek V3模型的区别是什么？什么时候该使用推理大模型？

API pricing

API price comparison

Side-by-side input/output token pricing

Higher is usually better; “—” means no score.