I tested DeepSeek Harness for a week. I left with a shipped plugin and $0 in API costs

Description and use

The dsh‑model‑pricing plugin added a pricing table to DeepSeek Harness, showing per‑million‑token costs and cross‑provider comparisons, all built and run with zero LLM cost.

Problem

Users of DeepSeek Harness lacked a way to see per‑model pricing and compare providers directly within the Settings UI.

Outcome

Implemented a fully functional pricing board inside DeepSeek Harness with zero LLM cost, enabling users to compare model prices and providers directly within the UI.

Workflow

• 1. Install the plugin with npx -y @deepseek-ai/dsh plugin --profile web add dsh-model-pricing
• 2. Open the harness Settings page and navigate to the Models tab
• 3. View the pricing board with per‑token costs, provider flags, and cheapest badges
• 4. Use the rule engine to filter by capability tags or subscription plans
• 5. Review the community promotions feed for free or discounted models

Evidence

Total LLM spend for the project: $0.

Source

per-1M-token input / output / cache prices for 7,178 priced models across 213 providers

Source

The harness's own routing catalog overlaid for the 788 models it can actually invoke

Source