Priced per report answered, not per seat.
Reporters never count as users — they don’t have accounts. You pay for the reports Command takes off your hands.
Solo
One product, one person answering.
100 reports a month
- Unlimited reporters
- Linear + GitHub
- Agent handoff, bring your own key
- Email support
Team — MOST TEAMS
A product team with real users reporting.
1,000 reports a month
- Everything in Solo
- EM triage agent
- CodeRabbit rework routing
- Silent-report escalation
- Shared Slack channel
Scale
Several products, several repos, an SLA.
10,000 reports a month
- Everything in Team
- Multiple repos and projects
- SSO, audit log, custom retention
- Self-hosted agent runners
- Named engineer on call
What’s in each plan
ALL PLANS · 30 DAYS FREEScroll to compare all plans
| FEATURE | SOLO | TEAM | SCALE | ENTERPRISE |
|---|---|---|---|---|
| Widget, unlimited reporters | Included | Included | Included | Included |
| Classification, scope, acceptance criteria | Included | Included | Included | Included |
| EM triage agent | Not included | Included | Included | Included |
| Silent-report escalation | Not included | Included | Included | Included |
| Repos connected | 1 | 3 | Unlimited | Unlimited |
| SSO and audit log | Not included | Not included | Included | Included |
| VPC or on-prem | Not included | Not included | Not included | Included |
Pricing questions.
What counts as a report?
One thing a person told you, however long the thread runs. Follow-up questions, her answers, a reopen — all the same report. Duplicates merged into an existing thread don’t count again.
Do reporters need seats?
No, and they never will. Reporters have no accounts — the thread lives in the widget in your app. Seats only apply to your own team in the back office, and those are unlimited on every plan.
What happens if we go over?
Nothing breaks and nothing gets dropped — every report is still received and answered. We bill the overage at the plan’s per-report rate and tell you before it happens.
Do we pay for the coding agents?
You bring your own keys for Codex, Cursor, Claude Code or opencode, and pay those providers directly. Command never marks up agent usage. Running with no agents at all costs the same.