A fast terminal dashboard for tracking token usage and costs across Claude Code, OpenAI Codex, Google Gemini CLI, Kimi Code, and Oh My Pi. Written in Rust.
- Multi-vendor usage tracking in one place
- Per-model token breakdown with cost analysis and daily/weekly/monthly projections
- ASCII time series charts with ANSI colors
- Responsive display adapting to terminal width
- Live monitor mode with interactive keyboard controls and process CPU/RSS telemetry
- Runtime monthly-cost editing with validated
.fee.envpersistence - Versioned JSON snapshots for local integrations
sh -c "$(curl -fsSL https://raw.githubusercontent.com/SihaoLiu/ai-usage/main/scripts/install.sh)"The installer detects Linux x86_64, Linux aarch64, or Apple Silicon; downloads
the matching latest release; installs it to ~/.local/bin/ai-usage; and
starts the dashboard immediately. It prints a PATH reminder when needed.
Set AI_USAGE_INSTALL_DIR to install elsewhere, or AI_USAGE_VERSION to a
specific tag such as v3.0.0. Once installed, type update in monitor mode
or use --auto-update for periodic self-updates.
# Standard build
cargo build --release
# Static binary (portable across x86_64 Linux)
rustup target add x86_64-unknown-linux-musl
cargo build --release --target x86_64-unknown-linux-musl# All vendors, last 3 days (monitor mode)
ai-usage
# Single snapshot, then exit
ai-usage --once
# Last 30 days, specific tool
ai-usage --days 30 --tool claude
# Machine-readable usage for the last 7 local calendar days
ai-usage snapshot --days 7 --tool allSee docs/usage.md for full usage details, data sources, monitor mode controls, and configuration. For optional cross-machine sync, see docs/sync.md.
MIT License - see LICENSE file.