Theme
AI Resources
ccusage
ccusage is an MIT-licensed command-line tool for reading local coding-agent usage logs and turning them into reports for tokens, models, sessions, and estimated costs. It can combine supported sources in one view or focus on a single tool such as Codex or Claude Code.
The same daily, weekly, monthly, and session views work across supported coding CLIs, with date filters plus structured JSON output for spreadsheets, scripts, or dashboards. Claude Code also has billing-block views and project grouping, while individual sources can expose their own focused options. Use this as a first read, not a recommendation. Open the original project before trusting details like terms, limits, privacy, cost, setup, or safety.
What it is
A usage report for coding-agent CLIs
It reads supported local usage files and groups token counts, models, cache activity, and cost data by day, week, month, or session. Focused commands can narrow the same report to one source.
Why it stands out
One view across several agent tools
A unified report can place Codex, Claude Code, Gemini CLI, OpenCode, and other detected sources in the same table. That makes it easier to see where usage is happening without opening each tool separately.
Availability
Try it from a package runner or Nix
Readers can run ccusage with npx, bunx, pnpm, or Nix, then use table output in a terminal or switch to JSON for further analysis. Offline mode uses cached pricing data when a network-free report is preferred.
Why it matters
What makes it useful
Coding-agent use can spread across several CLIs and local sessions. ccusage turns those separate logs into a shared timeline, so a reader can spot heavy days, compare tools, check model mixes, or export the numbers before deciding where a workflow is becoming expensive.
What to know
Where it fits
It fits individual users and practical builders who already use one or more supported coding CLIs and want a quick local report without building their own log parser. JSON output also gives people a starting point for a personal dashboard or a recurring usage check.
Notable points
What stands out
Treat the cost columns as usage estimates rather than an account bill. Pricing can come from stored cost fields or model-rate calculations, and the Codex adapter describes its support as experimental while Codex log formats continue to change. Codex cost output is an API-equivalent estimate, not a ChatGPT credit balance.
Before using
What to review
Check which local folders ccusage will read, especially when custom Claude or Codex homes include archived sessions or saved JSONL output.
Choose the cost mode deliberately when comparing periods. Auto, calculate, and display modes can use different inputs and may not produce the same totals.
Review JSON before sharing or publishing it. Reports can include dates, project or session groupings, model names, token totals, and estimated costs.
Use the source-specific documentation when a report looks incomplete. Older logs, missing model metadata, or changed formats can affect what is counted.
Use offline mode when cached pricing is sufficient and the report should avoid a pricing-data refresh.
Reader fit
Who may find it relevant
People switching between several coding-agent CLIs who want one usage view.
Codex or Claude Code users checking daily, monthly, session, or model patterns, with project grouping available for Claude Code.
Builders who want JSON output for their own spreadsheet, script, or dashboard.
Less useful for readers who need an official provider bill, account quota, or guaranteed historical reconstruction from incomplete logs.
Editorial note
Why LifeHubber lists it
LifeHubber lists ccusage because it turns scattered coding-agent logs into one readable usage history. That helps readers see which tools and models account for their token use and estimated costs, while keeping the difference between a local estimate and an official account balance clear.
Source links
Source materials
Reader note
Before relying on this entry
LifeHubber lists entries to help readers inspect AI projects, not to endorse them or prove they are safe, suitable, accurate, maintained, or right for a specific use. We do not verify every entry in depth. Before relying on anything listed, review the original materials, terms, privacy practices, limits, and risks that matter for your situation.
More in Productivity
Keep browsing this category
Explore more productivity tools.
JCode
1jehuang/jcode
A Rust-based terminal tool for running coding agents, with persistent memory, multi-session and swarm workflows, provider switching, local-model paths, reusable skills, MCP support, and cross-platform releases.
Meetily
Zackriya-Solutions/meetily
A local-first AI meeting assistant with public source, desktop release assets for Windows and macOS, Linux source-build docs, live transcription, meeting summaries, local Whisper or Parakeet transcription, local or external LLM summary options, optional analytics, and an MIT license.
Penpot
penpot/penpot
An open-source collaborative design platform with an official MCP server that lets AI agents inspect and modify Penpot files, components, tokens, styles, layouts, and assets through hosted or local connection paths.
