Unverified50% confidenceFactExact time
Perplexity本身不训练基础模型，而是通过调用OpenAI、Anthropic等公司的模型API来提供服务
1
Sources
50%
Confidence
Long-term
Relevance
8/8/2026
First Seen
Sources
Related Claims
VerifiedOpenAI has adopted an open-weight model rather than open source in the strict sense, without disclosing training data or training code.74% similarUnverifiedAltara Tech's approach is to leverage the reasoning capabilities of foundation models such as OpenAI, layering domain knowledge and workflow design to achieve vertical applications, rather than training domain models from scratch.74% similarVerifiedMajor service providers such as OpenAI and Anthropic generally commit to not using API data for model training.72% similarVerified开放权重模型仅发布训练完成的模型参数文件，而严格意义的开源AI要求公开训练代码、数据集与完整复现流程71% similarUnverified如果业务能直接调用OpenAI等API且成本可接受，则没有必要额外投入Agent Tuning训练资源70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/705714API
curl https://kongchang.com/api/v1/knowledge/claims/705714MCP
get_claim(id=705714)