Expired60% confidenceFactExact time
Anthropic adopts a tiered model strategy of Haiku/Sonnet/Opus, while OpenAI has the o1-mini and o1-preview tiers.
2
Sources
60%
Confidence
Medium-term (~90 days)
Relevance
7/11/2026
First Seen
Valid until: 10/9/2026(expired)
Sources
Related Claims
UnverifiedThe limited preview mode continues OpenAI's gradual release strategy used for reasoning models such as o1 and o3.79% similarVerifiedOpenAI's o1 series represents a new paradigm called "test-time compute"75% similarUnverifiedOpenAI已有mini/turbo系列，Anthropic有Haiku/Sonnet/Opus三档划分，体现用差异化模型规格覆盖不同成本-性能区间的行业共识74% similarVerifiedOpenAI的o1系列模型和DeepSeek-R1采用了类似Chain-of-Thought的显式推理机制73% similarUnverifiedOpenAI的o系列模型（o1、o3）是强化学习推理（RLR）路线的代表性产物，以可验证的客观标准作为奖励信号73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/480956API
curl https://kongchang.com/api/v1/knowledge/claims/480956MCP
get_claim(id=480956)