Unverified50% confidenceOpinionExact time
The original author's testing feedback states that deepseek-harness performs adequately across the three dimensions of speed, quality, and cost, but lacks specific benchmark data.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/12/2026
First Seen
Valid until: 1/10/2027
Sources
Related Entities
Related Claims
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67% similarUnverifiedThe author acknowledges that the DeepSeek model is somewhat verbose and currently not as capable as some dedicated tools when it comes to looking up information in real time or repeatedly testing until success.67% similarUnverifiedDeepSeek在MATH、AIME等数学竞赛基准测试上表现突出66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1000630API
curl https://kongchang.com/api/v1/knowledge/claims/1000630MCP
get_claim(id=1000630)