Who this is for
Developers and engineering leads on Claude Code and Codex
Track token spend, model mix, streaks, and project-level breakdowns for Claude Code and Codex CLI in one dashboard. Compare Opus, Sonnet, Haiku, GPT-5.5, and Codex side by side.
Outcomes
- See combined Claude + Codex spend in one place
- Catch runaway Opus usage before the bill lands
- Per-project attribution so you can bill usage back to the right budget
Why track Claude Code and Codex together
Most developers end up using both — Claude Code for design and reasoning, Codex CLI for fast code edits — and neither official dashboard shows them in one view. That means you cannot compare cost-per-task or model mix across the two tools. Vibecode does.
How it integrates
Run the WolfAI sync script against your local Claude Code session folder and Codex CLI logs. The script posts aggregated token counts to the Vibecode backend — prompts and responses never leave your machine. The dashboard then renders daily charts, model breakdown, heatmaps, and a yearly wrapped recap.
What you can learn
Patterns: which days are highest-cost, which projects use Opus vs Sonnet, how your usage shifts as you move from design to implementation. Anomalies: a single accidental Opus loop can cost more than a month of normal use — Vibecode surfaces that immediately instead of on the next invoice.
Related products
Models typically involved
Frequently asked questions
How do I track Claude Code usage?
Use Vibecode at wolfai.dev/vibecode. Run the sync script against your local Claude Code session folder — it extracts token totals per model and uploads only aggregated counts, not prompt text.
Is there a free ccusage alternative that covers Codex too?
Yes. Vibecode covers Claude Code and OpenAI Codex CLI in a single dashboard, while ccusage is Claude-only. Both are free.