Unverified50% confidenceconcept_definitionExact time
The OpenAI-compatible endpoint is a lowest-common-denominator abstraction that unifies fields such as messages, max_tokens, and finish_reason, but cannot express vendors' proprietary capabilities (such as Ollama's think or Anthropic's extended thinking).
1
Sources
50%
Confidence
Long-term
Relevance
10/16/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedPerplexity provides a Custom Inference Endpoint setting under the Advanced tab, which by default points to the local Ollama service and is compatible with any OpenAI-standard API endpoint.72% similarUnverifiedBeing compatible with the OpenAI format only guarantees that the interface shape is consistent; it does not guarantee consistency in output style, function calling support, context length, or rate limiting strategy.72% similarUnverifiedagentgateway's Ollama provider goes through the OpenAI-compatible endpoint, which completely ignores Ollama's native "think": false parameter, causing reasoning models' chain of thought to consume the entire max_tokens budget, the content to return an empty string, finish_reason to be length, and the HTTP status code to still be 200.72% similarUnverifiedOpenAI-compatible interfaces usually cover only the core fields, and implementations of extended capabilities such as tools, response_format, and streaming events are not fully consistent across vendors; when encountering unsupported features, they may return a parameter error or silently ignore it.71% similarUnverifiedOpenAI在o1的技术说明中表示，出于对齐监控和避免用户被未经对齐的原始想法误导的考虑，不会向用户直接展示完整的原始思维链71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1047898API
curl https://kongchang.com/api/v1/knowledge/claims/1047898MCP
get_claim(id=1047898)