deepseek-v4.1VIDEO
8 个 V4.1 Flash agent 同跑：单台 8x V100 约 40-50 tok/s
一台二手 8x V100 服务器跑自研 Rust+CUDA 引擎，同时服务 8 个 V4.1 Flash agent，约 40-50 tok/s，输出成本约 $0.20-0.25/百万 token。
Aleksey Smolenchuk
@lxe9 天前收录原推发布于 2026.10.23
Aleksey Smolenchuk @lxe原推发布于 2026.10.23
8 agents running DeepSeek-V4.1-Flash at once at ~40-50tps, All 8 are served from one used 8x V100 server running a custom a Rust + CUDA engine. This is about $0.20–0.25 per million output tokens on a spot-rented marketplace box.
x.com/lxe/status/2113382630284755403采集于 2026.10.23 12:06