Strata 0.2.4 本地跑 V4.1 Flash：64GB 内存 + 12GB 显卡实测

Jay Zhou @jayzhoupro原推发布于 2026.10.25

在 X 上打开原推

DeepSeek V4.1-Flash now runs inside 64GB of system RAM on a gaming PC. Strata (MIT) ships it in its Terminal build from 0.2.4 and serves an OpenAI/Anthropic API on localhost, so Claude Code and Codex can point at it. What the repo commits to: 32GB of RAM plus VRAM is enough to start, 64GB runs every size. On a 12GB RTX 5070 with 64GB of RAM, Q2_0 answers at 94 tokens/s, IQ3_S at 53, and prompt intake runs ~2,650 tokens/s. Q2 and IQ3 are 2-3 bit weights, so the open question is whether it holds a repo together. what would you test it on first?

这条是引用着另一条写的：「实测：V4.1 Flash 用 Strata 在 64GB 内存上跑起来」