Unverified50% confidenceFactExact time

Reply Desk uses the onnx-community/gemma-3-1b-it-ONNX model, using 4-bit weights on WebGPU and 8-bit weights on CPU.

1

Sources

50%

Confidence

Medium-term (~90 days)

Relevance

10/15/2026

First Seen

Valid until: 1/13/2027

Sources

Reply Desk: A Review Reply Assistant for Small Businesses That Keeps Data on-Device

Related Entities

Reply Deskproduct Gemma 3 1Bmodel

Related Claims

Unverifiedx86-64架构使用RSP寄存器作为栈指针寄存器来管理栈61% similar UnverifiedPagedAttention enables the same GPU hardware to serve 2–4 times the number of concurrent requests.61% similar Unverified栈由 CPU 硬件寄存器（如 x86 架构的 RSP/RBP）直接管理，随函数调用自动压栈、随函数返回自动弹栈61% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/1038453

API

curl https://kongchang.com/api/v1/knowledge/claims/1038453

MCP

get_claim(id=1038453)