Buying snapshot
Scroll sideways to see all columns. With a keyboard, focus this table and use the arrow keys.
| Decision point | What the CCDash project states |
|---|---|
License | MIT ("Use it however you want") |
Deployment | Self-hosted: clone the repository, run python3 server.py, open http://localhost:8420. No pip install, npm or Docker. |
Requirements | Python 3.8+ and an installed Claude Code CLI that has been used, so session data exists locally |
Data source | Scans the JSONL session files Claude Code writes to ~/.claude/projects/; calls the claude.ai API only if you add quota settings |
Release stage | v0.9.2, dated 2026-04-13; the roadmap lists v0.10, v0.11 and then v1.0 stable |
Platform notes | macOS launchd auto-start installer; Windows compatibility fixes shipped in v0.9.2 |
Our take
If you use Claude Code on a subscription and want to see your quota windows, model mix and API-rate cost on hardware you control, CCDash is worth a trial: it is free to run and takes a few commands to start. If you need a supported product with stable behaviour for finance reporting, treat it as a personal or small-team tool, not a system of record.
Evaluate CCDash
Comparisons
Where it stands out
No API key for usage data. The project describes itself as working with subscriptions without an API key, because it reads local session files. Engineers on Pro or Max plans can get token and cost views without new credentials.
Both quota windows, with a warning. HUD gauges cover the 5h and 7d quota, and a rate limit predictor shows a risk level, a time-to-throttle countdown and burn rates over 5, 15, 30 and 60 minutes. You see a limit coming instead of finding it mid-task.
Cost views you can act on. Per-call cost at official pricing is broken down per model and per project, with cache efficiency grading and suggestions such as Opus to Sonnet savings estimates. Git integration adds AI cost per commit. The overview also tracks average TTFT and response duration, and deep analytics show context window usage %.
Session detail. Each session opens a full conversation timeline with tool call badges and file operations, and every row shows a Working, Idle or Done status for local and remote sessions. This helps you check whether a long agent run has finished.
Alerts and terminal access. Daily, weekly and monthly token budgets show progress bars with thresholds at 60%, 80% and 100%. Webhooks to Slack, Discord or any HTTP endpoint fire when 5h quota use passes 80% or daily cost exceeds the budget, checked every 5 minutes. A CLI offers five commands (status, top, models, budget, live), and data exports to CSV or JSON.
More than one machine. A read-only agent (agent.py) on a remote host exposes the same file scanning over HTTP with a token, and the dashboard combines remote and local sessions. Multiple Claude accounts, such as personal and work, get their own quota tracking.
Product screenshot
Trade-offs to weigh
Quota gauges need a browser cookie. To unlock real-time quota gauges you copy the sessionKey cookie from claude.ai DevTools and your org ID into config.json. The key lasts weeks, so expect to replace it; CCDash shows a warning when it expires. Treat that file as a secret.
Quota data uses an unofficial route. The tech stack lists a Swift quota fetcher for macOS that "bypasses Cloudflare for claude.ai API". Because the project is not affiliated with Anthropic, this part may break if claude.ai changes.
Costs are estimates. CCDash calculates per-call cost from token usage at official pricing. That shows what usage would cost at API rates, not what your invoice says.
Remote coverage is manual. Each remote host needs the repository cloned and agent.py running, and the README's setup uses an SSH tunnel plus a token in your local config. A host without the agent is missing from the combined view.
Pre-1.0 software. With v1.0 stable still on the roadmap, plan for behaviour changes between releases before you build scripts on the CLI or exports.
Pricing and plan conditions
CCDash is MIT-licensed code that you clone and run; the project states there is no signup. What you pay for is your own Claude usage and the machine that runs the Python server, plus any remote hosts running the agent. The budget figures in config.json (for example daily 20, weekly 100, monthly 400 in the template) are your own cost limits in USD for Claude usage, not a CCDash charge.
Public signals
About these signals
Verified factual signals from public sources. They indicate observable activity or interest, not total adoption, product quality, or cost.
0 GitHub commits 90d73 GitHub stars
See all signals from 2 sources
Alternatives to CCDash
The reviewed substitutes for CCDash among the claude code tooling, and what would make each one the better answer.
Direct alternatives
Reviewed substitutes: products bought for the same job, where a team picks one.
- Claude Usage Tracker
- Both read Claude session data off the local disk and show what it cost, and a developer installs one of them. The 2026-09-10 rejection under IDENTITY-007 read CCDash as a session monitor against Claude Usage Tracker's cost tracker -- two different jobs. That was true of CCDash's March launch and is no longer true of the product: github.com/zihenghe04/CCDash now leads with 'track tokens, quota, costs across Claude Code, claude.ai and the API in one panel' and has dropped task scheduling. The split is breadth against depth. Claude Usage Tracker reads 10+ tools -- Cursor, Windsurf, Cline, Aider, Continue.dev, Codex -- and reports historical cost. CCDash reads only Claude surfaces but adds what the other does not have: 5-hour and 7-day subscription quota tracking, burn-rate and rate-limit prediction, budget thresholds and Slack or Discord alerts.Applies to: Seeing what AI coding assistants cost, from local session files, without an API key. Claude Usage Tracker stands in when the spend is spread across several editors and agents and the question is what was spent. CCDash stands in when the work is Claude-only and the question is how close the next throttle is.
Related technologies
Normally used together rather than chosen between, so these are not alternatives.
- capsul
- Both target Claude usage limits from different sides. CCDash reads Claude Code's local session files and tracks the 5-hour and 7-day quotas, burn rate and rate-limit prediction; capsul lowers what each request sends. Watching quota does not reduce a request, and sending less does not show how close the next throttle is, so a developer hitting limits can use both.Applies to: Hitting Claude subscription limits: watch quota and burn rate with CCDash, lower per-request context with capsul.