clanker-analytics 0.2.1

AI coding tool token analytics powered by DuckDB

clanker-analytics

Token usage analytics for AI coding tools. Reads local session logs and shows per-project breakdowns using DuckDB.

Supports Claude Code, Codex, and Gemini CLI.

Install

uv tool install clanker-analytics

Or run without installing:

uvx clanker-analytics

Usage

clanker-analytics                        # ranked by project (default)
clanker-analytics --since 24h            # last 24 hours (also: 7d, 2w, 2026-03-01)
clanker-analytics --by date              # daily totals
clanker-analytics --by model             # per-model breakdown
clanker-analytics --by session           # per-session breakdown
clanker-analytics --tool claude          # Claude Code only
clanker-analytics --tool codex           # Codex only
clanker-analytics --tool gemini          # Gemini CLI only
clanker-analytics --chart                # generate PNG chart
clanker-analytics --share                # generate chart + open X compose
clanker-analytics --refresh              # force cache rebuild
clanker-analytics --sql "SELECT ..."     # custom SQL against 'tokens' table

Table view

Use --table for a tabular breakdown:

$ clanker-analytics --table --since 7d --limit 10
┌─────────────┬─────────────┬───────┬─────────┬──────────┬─────────┬─────────┬──────────┬────────────┬────────────┐
│   project   │    tool     │ turns │  total  │ billable │ output  │  cache  │ api_cost │ first_seen │ last_seen  │
├─────────────┼─────────────┼───────┼─────────┼──────────┼─────────┼─────────┼──────────┼────────────┼────────────┤
│ *           │ *           │ 60060 │    9.1B │    1.2B  │   18.3M │  97%    │ $9802    │ 2026-03-10 │ 2026-03-17 │
│ *           │ Codex       │ 37129 │    5.5B │  753.1M  │   13.8M │  96%    │ $7662    │ 2026-03-10 │ 2026-03-17 │
│ *           │ Claude Code │ 20303 │    3.6B │  417.9M  │    4.3M │  98%    │ $2089    │ 2026-03-10 │ 2026-03-17 │
│ *           │ Gemini      │  2628 │   36.4M │   18.2M  │  205.2k │  56%    │ $50.7    │ 2026-03-10 │ 2026-03-17 │
│ tao-auto    │ Codex       │ 12587 │    1.9B │  250.9M  │    4.1M │  97%    │ $2694    │ 2026-03-10 │ 2026-03-16 │
│ tao         │ Codex       │ 10670 │    1.8B │  245.0M  │    4.5M │  96%    │ $2470    │ 2026-03-10 │ 2026-03-16 │
│ wezterm     │ Claude Code │  2237 │    1.2B │  132.5M  │  410.3k │  99%    │ $692     │ 2026-03-13 │ 2026-03-15 │
│ wildlands   │ Codex       │  5515 │  714.8M │   94.8M  │    1.7M │  97%    │ $994     │ 2026-03-14 │ 2026-03-16 │
│ tao-perf    │ Codex       │  4960 │  673.0M │   86.4M  │    1.8M │  97%    │ $939     │ 2026-03-10 │ 2026-03-12 │
│ tao-sprites │ Claude Code │  5954 │  609.5M │   72.1M  │    1.4M │  98%    │ $247     │ 2026-03-10 │ 2026-03-16 │
└─────────────┴─────────────┴───────┴─────────┴──────────┴─────────┴─────────┴──────────┴────────────┴────────────┘

How it works

DuckDB reads session logs directly from ~/.claude/projects/, ~/.codex/sessions/, and ~/.gemini/tmp/ — no Python JSON parsing. Results are cached to ~/.cache/clanker-analytics/tokens.parquet (ZSTD compressed) and auto-invalidated when source files change.

Columns

• total — all tokens processed (input + output + cache write + cache read)
• billable — total minus the 90% cache read discount
• output — output tokens only
• cache — cache read hits as % of input tokens
• api_cost — estimated cost at API rates

API cost calculation

The api_cost and billable columns use published API pricing. Cache reads are 0.1x the input token price for all three providers:

	Input	Cache read	Cache write	Output
Claude Sonnet	$3/MTok	$0.30/MTok	$3.75/MTok	$15/MTok
Claude Opus	$5/MTok	$0.50/MTok	$6.25/MTok	$25/MTok
GPT-5	$1.25/MTok	$0.125/MTok	(auto)	$10/MTok
Gemini 2.5 Pro	$1.25/MTok	$0.315/MTok	(auto)	$10/MTok

Sources: Anthropic pricing, OpenAI pricing, Google AI pricing

Environmental impact estimates

The --chart / --share output shows estimated environmental impact per million tokens:

Metric	Per 1M tokens	Source
Electricity	0.6 kWh	Epoch AI, arxiv:2505.09598
Water	1 liter	Li & Ren (2023), adjusted for modern models
CO2	90 g	Ritchie (2025)

These are rough estimates — actual impact varies 10-100x depending on model, hardware, and data center location. No provider publishes official per-token figures.

Chart colors

Brand colors used in --chart / --share output:

Tool	Color	Source
Claude Code	#d97757	Anthropic brand guidelines
Codex	#10a37f	OpenAI brand
Gemini	#4285f4	Google brand

Requirements

Python 3.13+, DuckDB 1.5+, matplotlib 3.9+.

Tested on Linux, macOS, and Windows (including WSL data auto-discovery).