dsh-cost-display

You have been running models in DeepSeek Harness for hours, token counters climbing steadily, yet you have no idea what the session is actually costing. dsh-cost-display closes that gap by appending a real-time cumulative cost (Cost ¥X.XX) right after the token stats in the status bar. Hovering over it reveals a per-model breakdown that also separates reasoning tokens, so every dollar spent is transparently accounted for.

The pricing engine goes well beyond a flat rate. Peak and off-peak windows—defaulting to 09:00–12:00 and 14:00–18:00 Beijing time—let the plugin automatically select the correct tier at request time. Each model supports multiple price tiers with effective dates, so an official rate change is simply a new row in the table, no code edits required. When you add or rename a model in the settings page, the plugin syncs it into the price table on its own; unlisted models fall back to a safety rate marked with a ? prefix, and per-model peak pricing can be toggled off individually.

If you rely on DeepSeek for long-form writing, code generation, or heavy reasoning tasks and want a quiet, always-updating cost tracker beside your work without breaking your flow—or if you batch jobs during off-peak hours to cut unit prices—this lightweight plugin gives you a clear, real-time, and fully configurable picture of what you are actually spending.