Success-Only Billing
You're only charged for conversions that complete successfully. A file that fails, times out, or never gets attempted because a batch was cancelled or interrupted does not count against your monthly conversion allowance.
Billed
- A file that completes the pipeline and produces output, regardless of its final confidence grade
Not Billed
- A file that fails discovery, planning, conversion, or times out
- A file that was never attempted because you cancelled or paused before it ran
- A failed file that gets retried on resume, whether or not the retry succeeds counts only once, on its own outcome
What's Billable
The unit that counts against your plan is a successfully converted file. Every attempt (success or failure) is recorded for your own visibility in session info and the batch report, but only the successful ones are deducted from your monthly allowance.
| Plan | Conversions / Month |
|---|---|
| Free | 5 |
| Pro | 200 |
| Team | 1,000 |
| Enterprise | Unlimited |
Run 3xcode status at any time to see your plan, conversions used this month, and remaining allowance.
Quota Held for a Running Batch
Starting a batch puts a quota hold on the number of files it contains, up front, so a large batch can't accidentally overrun your remaining allowance partway through. As files finish successfully, they're billed against that hold; when the batch finishes, is paused, or is cancelled, whatever part of the hold is still unused is released back to your account immediately, not at the end of the billing period.
Held at the start
Creating and running a batch places a hold covering its file count, so your remaining quota reflects work already in flight, not just work already billed.
Released when unused
Finish, pause, or cancel a batch and any part of the hold that wasn't consumed by successful conversions is released right away.
# Check your current plan, usage, and remaining allowance 3xcode status # A cancelled batch releases its unused hold immediately 3xcode batch cancel <session_id> <batch_id>
Resuming a batch that has failed or unfinished files re-runs only those files, still under the same success-only rule: a file that failed once and succeeds on retry is billed exactly once, for the successful attempt.
Confidence Bands
Every successfully converted file gets a confidence score from 0-100, and every batch report groups files into four bands so you can triage a large batch at a glance instead of reading every audit file in file order.
| Band | Range | What it means |
|---|---|---|
| Perfect | 100 | No validation issues found. Still worth a spot check, not a substitute for one. |
| High | 90-99.99 | Minor issues at most, likely auto-fixed. Safe to review quickly rather than line by line. |
| Medium | 70-89.99 | Real issues found. Read the audit's manual-review items before treating the file as done. |
| Low | Below 70 | Significant issues, or constructs with no safe automatic fix. Prioritize these first. |
Billing does not depend on the band: a Low-confidence file that completed the pipeline is billed the same as a Perfect one, since both are successful conversions. Bands are a triage tool for your review, not a billing tier. See Batch Conversion & Sessions for how bands show up in a batch report.
Org-Shared Quota
Individual, per-account quota is the default for everyone, org membership doesn't take it away. On top of that, an organization can be given a shared quota pool by an admin. When one exists, quota resolution is org-first: conversions draw against the shared pool (or a per-seat allotment within it, depending on how the admin configured it) before ever touching your individual quota.
Check which quota you're drawing from
3xcode org me shows whether your org has a shared quota and its mode: per-seat or shared pool. 3xcode status always shows your own individual usage regardless.
Individual quota as fallback
If there's no org, no shared quota, or the shared quota is exhausted, conversions fall back to your individual plan quota automatically.
See Organizations & Licensing for how to create or join an organization and how a shared quota gets provisioned.