Unverified50% confidenceBenchmarkExact time
Running the GLiNER 2.5 model on MPS, the average p50 latency is 21.1 milliseconds.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/21/2026
First Seen
Valid until: 11/4/2026
Sources
Related Entities
Related Claims
UnverifiedSnapshot restoration time is now counted in the creation phase, and creation time has increased by about 50 milliseconds at p50; for uncached snapshots, the increase may be larger74% similarUnverifiedIn Microsoft's internal testing, Decision-1's P50 latency is approximately 85 milliseconds.73% similarUnverifiedIn LLM scenarios, latency hedging is the most commonly used strategy; in practice, the P50 or P75 latency value is typically used as the trigger threshold, and hedging is triggered if the first request has not returned its first token within the P75 time.72% similarUnverified中继转发相比P2P直连往往会引入额外的30-80ms延迟70% similarVerifiedIn testing, the p99 latency per 160-millisecond tick was 147.4–147.5 milliseconds, with zero missed deadlines across three rounds totaling 84,000 session ticks measured.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1081946API
curl https://kongchang.com/api/v1/knowledge/claims/1081946MCP
get_claim(id=1081946)