Unverified50% confidenceBenchmarkExact time
DeepSeek-R1 approaches or matches GPT-4-level performance on standardized benchmarks such as MATH and HumanEval.
1
Sources
50%
Confidence
Long-term
Relevance
9/30/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedDeepSeek-R1 distilled models perform strongly on tasks such as AIME 2024, MATH-500, and LiveCodeBench, matching GPT-4o and DeepSeek V3 in most scenarios and surpassing OpenAI o1-mini on some metrics.76% similarUnverifiedIn early 2025, DeepSeek-R1 achieved performance on par with OpenAI o1 on multiple international standard benchmarks including the AIME mathematics competition and MMLU, with significantly lower training costs than GPT-4.75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/962579API
curl https://kongchang.com/api/v1/knowledge/claims/962579MCP
get_claim(id=962579)