Free
$0
For trying it out
- 1 workspace
- 3 signal sources
- 50 loop cycles / month
- 7-day memory retention
- Approval-first queue
- All integrations
- Public API v1 access
- Sources
- 3 sources
- Memory
- 7-day TTL
- Loop interval
- 15-minute floor
Pricing
Flat monthly fee. No credits. No surprises.
Loop Desk charges a flat monthly subscription — unlimited cycles, no per-cycle credits, no metered AI tokens. Every tier ships every integration, every API endpoint, and the same approval-first guarantee. Higher tiers unlock higher hard caps (sources, memory retention, cycle cadence) — they do not unlock different products.
No per-cycle credits No metered AI tokens No overage bills Daily cost cap on every tier
Pro
$29/mo
For operators who ship
- 5 workspaces
- 25 signal sources
- Unlimited loop cycles
- 90-day memory retention
- Higher daily cost cap
- Priority cycle queue
- Full integrations + API v1
- No per-cycle credits, ever
- Sources
- 25 sources
- Memory
- 90-day TTL
- Loop interval
- 5-minute floor
Team
$79/mo
For growing teams
- 25 workspaces
- 100 signal sources
- Unlimited cycles + memory
- Multi-operator + role-based perms
- Dedicated support
- Highest daily cost cap
- Full audit trail + CSV exports
- SOC 2 / ISO 42001 evidence ready
- Sources
- 100 sources
- Memory
- No TTL (permanent)
- Loop interval
- 1-minute floor
How we compare
Why flat-fee, not per-cycle credits
The AI-tooling market in 2026 has split. Notion went per-cycle credits in May (you pay each time the agent runs). HubSpot Breeze went outcome-based (you pay per resolved conversation or qualified lead). Loop Desk stays flat because budget anxiety is the #1 reason owner-led teams stall on AI adoption (McKinsey, March 2026 — 53% prefer mostly-human-led AI; IDC 2026 — variable billing is the top SMB blocker).
ToolBilling modelLoop Desk approach
Notion Custom AgentsPer-cycle credits. $10 / 1,000 credits (May 2026). Each cycle consumes credits scaling with output complexity.Flat monthly fee. Unlimited cycles on Pro and Team.
HubSpot BreezeOutcome-based. $0.50 per resolved conversation (Customer Agent). $1 per qualified lead (Prospecting Agent).Flat fee. No charge per output, per signal, or per approval.
Per-token billing (generic)Pass-through model bills with markup. Your bill grows with your activity AND with model price changes.Daily cost cap (rev 20) is a hard ceiling. Auto-pauses at cap, regardless of model bill drift.
Procurement evidence
Every tier ships the full product
Higher tiers unlock higher hard caps — sources, memory retention, loop cadence floor. They do not unlock different integrations, a different approval queue, or different governance controls. The procurement story is consistent at every tier: every output goes through the approval boundary, every cycle writes the activity log, every memory entry stamps its kind + importance.
Approval-first on every tier
No output publishes without explicit human sign-off. Free, Pro, and Team all queue every brief / draft / decision / watchlist / note for approval. No premium-tier auto-publish.
Audit trail on every tier
Activity log, per-cycle cost, per-task work log, version chain on artifacts, and JSON + CSV exports are available on every tier. SOC 2 / ISO 42001 evidence is consistent regardless of plan.
Daily cost cap on every tier
The hard daily spend ceiling (rev 20) is enabled on every tier. Auto-pauses the desk if today's OpenAI spend reaches the cap. Free defaults conservatively; Pro and Team raise the ceiling.
What we don't do
No per-cycle credits. No metered AI tokens. No paywalled integrations. No paywalled API endpoints. Higher tiers unlock more of what every tier already has.
Start free. Upgrade when your team needs higher caps.
One workspace, three sources, fifty cycles a month, the full integration set, the full API. Enough to verify Loop Desk fits your workflow before paying for higher caps.