Every number in a row belongs to one engine: its own cards, context, experts, speed and CPU. Speeds, load and power carry the minute of history the top cards used to show together, each metric in its own color; the experts and the context show as fills instead. RAM and disk belong to the whole computer, not to one engine, so they sit below; the CPU load and the board temperatures (CPU, memory, SSD) are the computer's too and no engine's, so they stay as cards; PCIe throughput is NVIDIA only (AMD cards do not report it, and Windows AMD cards report no per-card readings at all).
| Time | Status | Replica | Prompt | Reused | Output | Prefill t/s | Decode t/s | VRAM hit rate | Duration |
|---|---|---|---|---|---|---|---|---|---|
| No requests yet | |||||||||
Their tools run on this PC with your rights, when the model decides to call them (in this page's chat only; switch it off in Sampling).
Any OpenAI- or Anthropic-compatible client works with these addresses.
Chats, settings and the key are kept in this browser only. Strata on GitHub
Kept in the model's run config for every client. An empty field is the default. They take effect the next time the model starts (close Strata's window and start it again).