Highlights
- Group usage by model instead of date
- Useful for stack reviews and experimentation
- Pairs well with wrapped summaries and cost views
Feature
Model breakdown
Inspect which Claude-compatible or Codex models dominated your usage and how their token share changed over time.
Intent
Feature
Focus
Model breakdown in vibestats is designed for developers who want to know which models they actually relied on rather than which ones they planned to use.
Next step
Relevant commands
npx vibestats --modelnpx vibestats --claudenpx vibestats --minimaxWhat model breakdown means
Model breakdown is about getting a clear, local view of AI coding activity without building custom spreadsheets or one-off scripts.
Who this is for
Developers who want to know which models they actually relied on rather than which ones they planned to use.
How vibestats handles it
vibestats supports model-oriented reporting so you can view totals by model family, compare usage share, and see how the same date range looks when grouped by model instead of time.
What you get back
You get a cleaner answer to whether a specific model family drove most of your work, costs, or output across a reporting window.
FAQ
Can vibestats help with model breakdown?
Yes. vibestats is built to support model breakdown with local reporting, aggregate pages, and command-line workflows.
Does this stay local?
Yes. vibestats is local-first. Prompts and responses stay out of the hosted pages unless you intentionally publish aggregate share output.
Where should I start?
Start with the command reference, then use the related links below to jump into wrapped reports, activity heatmaps, share pages, or diagnostics.
Related pages
Continue by intent
guides how-to-compare-model-usageOpen the related vibestats page and continue the workflow from there.compare session-breakdown-vs-model-breakdownOpen the related vibestats page and continue the workflow from there.features combined-ai-coding-statsOpen the related vibestats page and continue the workflow from there.