AI Agent Hub
Back to plugins
🖥️

deepseek-cost-usage-status-plugin

Client Updated 2026.08.26

Run the following command in DeepSeek Harness:

dsh plugin install Zenjibad/deepseek-cost-usage-status-plugin

Paste the following prompt into your AI chat to install this plugin:

Run dsh plugin install Zenjibad/deepseek-cost-usage-status-plugin in your terminal to install the plugin; the source repository is at https://github.com/Zenjibad/deepseek-cost-usage-status-plugin. Restart DSH after installation to activate it.

About this plugin

When you are deep in a DSH conversation with DeepSeek, there is little visibility into exactly how much that session has cost, what your account balance looks like, or whether you are currently in a peak-billing window. deepseek-cost-usage-status-plugin fills that gap by rendering a compact, colour-coded status line right beneath the native conversation stats, surfacing live cost, token usage, burn rate, account balance, and the active model at a glance.

The line leverages DeepSeek's official Beijing-time peak windows (09:00-12:00 and 14:00-18:00) to show a green chip during off-peak (50% discount active) and a red chip during peak. Every streamed token is priced against the official CNY rate table, the 50% off-peak discount is applied automatically, and the resulting cost is converted into whatever currency your balance API returns (typically USD) using a live FX rate refreshed hourly, with a configurable fixed-rate fallback. Burn rate (cost per minute since the first model call), current model, and reasoning effort are displayed alongside, all in a 12px muted-tertiary style that reads as part of the same visual family as the shipped stats cell.

The plugin is ideal for DSH users who routinely call the DeepSeek API and want to stay on top of spend: checking that your balance will cover a long coding session, monitoring burn rate during batch jobs, or simply timing your work to off-peak hours to take advantage of the discount -- all without ever leaving the editor.

Use Cases

  • Check real-time session cost and account balance during long conversations to avoid running out of funds
  • Monitor per-minute burn rate during batch inference jobs to stay within budget
  • Time API calls to off-peak windows to automatically capture the 50% discount

Best For

  • Regular DSH users who call the DeepSeek API frequently
  • Developer teams that track token spend and account balance
  • Users running long-running or batch model inference who need fine-grained cost control