How tokens are used
What drives cost
- ✓Repository size (SLOC) and number of files analyzed
- ✓Enabled analyzers (languages/features)
- ✓Follow-up analysis passes (when needed for coverage)
- ✓Optional LLM report generation (answers + evidence appendix)
PViz shows an estimate before starting paid work. Actual usage is recorded on completion.
LLM report question limits
Guardrails for quality + predictability
- ✓Up to 5 questions per job (max)
- ✓Each question is limited to 350 characters
- ✓Pricing uses the parsed unique question count (post-dedupe, post-cap)
- ✓Email delivery includes answers + a link to the full report (no evidence blocks in email)
You’ll see what was accepted/dropped (duplicates, truncation, max limit) before submission.