dsh-conversation-cost
Run the following command in DeepSeek Harness:
dsh plugin install Uersx/dsh-conversation-cost
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install Uersx/dsh-conversation-cost in your terminal to install the plugin; the source repository is at https://github.com/Uersx/dsh-conversation-cost .
About this plugin
When you run long conversations with DeepSeek, the real cost often stays a mystery. Token usage is buried in session logs, peak and off-peak rates differ by 2x, and cache hits carry steep discounts — calculating it by hand is tedious and error-prone.
dsh-conversation-cost automates the whole thing. It reads the authoritative usage (input, output, cache read/write, reasoning tokens) returned by the provider in each assistant message, applies the official DeepSeek V4 pricing table, and buckets every call into peak (9:00-14:00, rate x2) or off-peak based on its actual timestamp. A persistent badge at the bottom of the sidebar shows the live cost and call count, refreshing every five seconds; the model can also invoke the conversation_cost tool to retrieve a per-call breakdown and a summary on demand.
It is built for developers and teams who work with DeepSeek V4 daily and want to keep an eye on per-session spend in real time, as well as anyone who needs to report inference costs to clients or stakeholders. All figures are estimates based on the official rate card; your platform bill remains the source of truth.
Use Cases
- Checking the total cost of a long conversation at a glance
- Splitting inference fees into peak and off-peak buckets for accurate accounting
- Reporting per-session cost breakdowns to clients or stakeholders
Best For
- Engineers who use DeepSeek V4 for daily development and debugging
- Team leads who need to track and report inference spend
- Product teams building DeepSeek-powered apps and need to surface costs to end users
Related Plugins
ModLens is a vision plugin for DeepSeek Harness that gives text-only models sight by reading images pasted directly into chat, with zero-config setup and multiple vision engines.
On-demand vision for text-only DeepSeek Harness agents: built-in free keyless vision chain and 14 vision tools, routing image turns as tool calls to vision models with pixel fidelity, no Python needed, one-command install.
Give text-only models in DeepSeek Harness eyes, enabling image Q&A, long-screenshot OCR, UI restoration, and GUI visual tasks.
An armor-breaking plugin for DeepSeek: rewrites instruction priority, expands refusal-blocked phrases, persists across sessions, and shows a green active indicator.