Unverified50% confidenceFactExact time
Jev treats all questions in the same request as a batch of parallel reasoning tasks, so twenty questions have almost the same latency as one question, and output tokens are not billed separately.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedJev can answer multiple questions in parallel within a single inference pass, making the time to answer 30 questions roughly the same as answering 1.72% similarUnverifiedMulti-query retrieval uses an LLM to automatically generate 3–5 semantically equivalent but differently phrased variant queries for the same question, retrieves results for each separately, and takes the union to improve recall.59% similarUnverifiedDecision Models support four question types (Choice, Null boolean, Score, etc.) and can be mixed within the same request; Questions can contain 1 to 64 named questions.58% similarVerifiedThe question decomposition strategy breaks complex questions into multiple sub-questions for separate retrieval and then consolidates the results, avoiding the inability to obtain sufficient information from a single complex query.58% similarVerifiedJev offers three question types: Select, Score, and Null/Bool, with each response accompanied by a confidence value.57% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1006157API
curl https://kongchang.com/api/v1/knowledge/claims/1006157MCP
get_claim(id=1006157)