AI Agent Hub
Back to plugins
🧰

dsh-llm-balance-siderbar

Web Tools Updated 2026.09.12

Run the following command in DeepSeek Harness:

dsh plugin install AlienGene/dsh-llm-balance-siderbar

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

Search for "usage" or "balance" in the DeepSeek Harness plugin market to install with one click, or run dsh plugin install AlienGene/dsh-llm-balance-siderbar in your terminal; source repository: https://github.com/AlienGene/dsh-llm-balance-siderbar

About this plugin

Running large models is smooth, but checking "how much quota is left this month" or "am I about to hit a hard limit" still means hopping between vendor dashboards. dsh-llm-balance-siderbar tucks all of that into a lightweight card in the corner of your DSH interface: pay-as-you-go accounts (DeepSeek, Moonshot) show a plain balance figure, while subscription accounts (Kimi For Coding, OpenCode Go) list per-window usage percentages alongside their reset countdown. Color coding does the judgement for you — red below 20% remaining, yellow below 40%, blue otherwise — and the provider you are currently using gets a small green dot.

The card is genuinely interactive: cycle through "current model / aggregate / all sources" views, drag the title to magnet-snap it flush against the left or right edge, collapse it into a single-letter bubble showing just one key number, fine-tune the refresh interval, or toggle a translucent background. Every position, fold state, and preference is remembered across sessions. No extra credential setup is needed; the plugin auto-detects whichever accounts you have already linked via API keys. Keys never leave the host process — the browser only ever receives amounts, currencies, window durations, and error codes over a read-only, non-caching API.

Built for developers and teams juggling two or more AI providers who want a constant, glanceable pulse on remaining balance and rate-limit usage without leaving their workflow.

Use Cases

  • Monitor balance and rate limits across multiple AI providers at a glance
  • Check remaining quota mid-session to decide whether to throttle requests
  • Get early color-coded warnings before subscription windows reset and usage is cut off

Best For

  • Developers calling three or more LLM APIs side by side
  • Engineering teams that need to track API spend and rate limits continuously
  • Individual developers making high-volume calls across multiple model providers