pi-deepseek-cost

简体中文

A Pi extension that tracks your DeepSeek usage & cost — live session cost in the status bar, a /ds-cost overlay panel, bilingual UI (中文/English) with currency switching, and official peak/off-peak billing.

Screenshots

Features

• Status bar: live cumulative cost for the current session (¥0.019), refreshed after every turn — cost only, no noise (an empty session shows ¥0)
• /ds-cost: floating overlay panel — per-model token usage (cache-hit / cache-miss input, output), cache hit rate, CNY/USD cost breakdown, peak-hour split, USD↔CNY cross-reference
• Model-aware: active only for DeepSeek models (provider: "deepseek" or id starting with deepseek); invisible otherwise
• Bilingual + currency: zh → ¥ (CNY), en → $ (USD). Switch inside the panel with L, or globally with Ctrl+Shift+L
• Official peak/off-peak pricing: intrinsic — each message is charged at the peak (×2) or off-peak rate of its own timestamp (Beijing 09:00–12:00 & 14:00–18:00)

Installation

# from npm
pi install npm:pi-deepseek-cost
# or from a git repo
pi install git:github.com/hacxy/pi-deepseek-cost
# or local path
pi install ./pi-deepseek-cost

Security: as with any Pi package, review the source before installing — extensions run with full system access.

Requires:

• A recent Pi installation
• DeepSeek models configured with an API key (DEEPSEEK_API_KEY, provider deepseek — models deepseek-v4-flash, deepseek-v4-pro)

Usage

Action	How
View session cost	Watch the status bar (updates after each turn)
Detailed cost panel	/ds-cost (Esc to close, L to switch language)
Switch language (zh↔en)	L inside a panel, or Ctrl+Shift+L (rebindable in keybindings.json)
Set language in config	deepseekCost.locale in settings.json

Configuration

In ~/.pi/agent/settings.json (global) or .pi/settings.json (project, overrides):

{
  "deepseekCost": {
    "locale": "zh"
  }
}

Key	Default	Description
locale	"zh"	UI language: "zh" or "en". Currency follows (zh → ¥, en → $)

Config edits take effect immediately (re-read on every calculation).

Peak/off-peak pricing is not configurable: DeepSeek bills peak ×2 during Beijing 09:00–12:00 and 14:00–18:00 (off-peak is exactly half of peak). Legacy peakPricing / peakMultiplier / peakHours keys in settings.json are silently ignored and can be removed.

Peak pricing details

Peak/off-peak pricing is official and always in effect. Each message is charged against its own timestamp: peak-hour messages at the peak rate (×2), off-peak at the off-peak rate — mixed sessions split precisely (the /ds-cost panel always shows off-peak / peak rows). Time is always evaluated in Asia/Shanghai (UTC+8, no DST), matching DeepSeek's official definition (09:00–12:00 and 14:00–18:00 Beijing time).

Cost basis

• Token numbers come from the real usage blocks Pi persists on messages (input = cache-miss input, cacheRead = cache-hit input, output), matching DeepSeek API's prompt_cache_hit_tokens / prompt_cache_miss_tokens / completion_tokens
• CNY and USD both use DeepSeek's official per-million-token prices as printed on the pricing page (≈¥6.9/$). Off-peak is exactly half of peak.

Model	Period	Input · cache miss	Input · cache hit	Output
deepseek-v4-flash	Off-peak	¥1.5 / M ($0.22)	¥0.05 / M ($0.007)	¥4.5 / M ($0.66)
	Peak ×2	¥3 / M ($0.44)	¥0.10 / M ($0.014)	¥9 / M ($1.32)
deepseek-v4-pro	Off-peak	¥4.5 / M ($0.66)	¥0.15 / M ($0.022)	¥13.5 / M ($1.98)
	Peak ×2	¥9 / M ($1.32)	¥0.30 / M ($0.044)	¥27 / M ($3.96)

deepseek-chat / deepseek-reasoner are deprecated (since 2026-07-24) and priced at the V4 Flash rates — kept so older persisted sessions still show correct costs.

• Cost is computed per model (model switches mid-session aggregate correctly); toolResult / compaction usage uses the last known model
• Session totals rebuild from the session file, so they stay accurate after /resume

Development

pnpm install
pnpm dev          # run in pi with hot reload (pi -e ./src/index.ts)
pnpm test         # vitest unit tests
pnpm typecheck
pnpm lint
pnpm format

Release

pnpm release [patch|minor|major]   # bump, push, and watch the CI publish (default: patch)

Bumps the version via npm version (commit + vX.Y.Z tag), pushes main + tags, then watches the publish.yml CI run: lint → typecheck → test → npm publish (npm trusted publishing — no tokens) → changelogithub GitHub Release. Use pnpm release --dry to preview the commands without running them.

License

MIT