Unverified50% confidenceBenchmarkExact time

DeepSeek-R1于2025年初在AIME数学竞赛、MMLU等多项国际标准基准测试中达到与OpenAI o1同级别的表现，且训练成本远低于GPT-4

1

Sources

50%

Confidence

Medium-term (~90 days)

Relevance

7/13/2026

First Seen

Valid until: 10/11/2026

Sources

GPT5.6免费用是陷阱？AI镜像站真相全揭露

bilibili小星说AI7/10/2026

Related Claims

UnverifiedDeepSeek-R1在发布时以远低于同类闭源模型的训练成本，在多个推理基准上与OpenAI o1持平83% similar UnverifiedDeepSeek以极低的训练成本实现了与顶级闭源模型接近的数学与代码能力，其R1模型在发布时引发全球关注78% similar Verified2025年初，DeepSeek-R1以约600万美元的训练成本达到了与OpenAI o1相当的推理能力78% similar UnverifiedDeepSeek-R1的推理能力在多个基准测试中表现惊艳，训练成本远低于国际竞品77% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/499524

API

curl https://kongchang.com/api/v1/knowledge/claims/499524

MCP

get_claim(id=499524)