Documentation

Limits and quotas

The caps that apply to Projects, Tasks and execution, and what occurs when you hit one.

Limits vary by Plan. This page outlines which limits exist and how they behave; the current figures for each Plan appear on the pricing page.

Projects

The free Plan permits a single Project. Paid Plans have no cap on Projects.

Agent slots

Slots determine how many Tasks run simultaneously, not how swiftly any Task finishes. The free Plan includes one low-priority slot; Professional offers one standard slot with extra capacity each month; Enterprise operates under a contracted arrangement with a minimum commitment and is in private preview.

When all slots are occupied, new Tasks queue in PENDING and begin once slots become available, ordered by priority. No Task is rejected due to lack of capacity – it simply waits.

See Capacity and agent slots.

Agent-hours

Plans with a monthly agent-hour allowance measure time while an agent is actively working. Queue time is complimentary.

Once the allowance is used up, new work ceases to start until the allowance resets or you switch to a Plan with greater capacity. Work already in progress is not terminated mid-Task.

Maximum task run time

A single Task may run for a set period before being halted. This ceiling is lowest on the free Plan.

Reaching this limit usually indicates the Task was too large rather than too slow. Split it – see Tasks, Plans, sprints and lanes.

Parallel subagents

The free Plan does not support parallel subagents. Paid Plans allow an agent to distribute work internally, which primarily impacts how swiftly it explores a large codebase.

Members

The free Plan is designed for one person trialling the product. Professional and Enterprise Plans have no cap on Members.

Adding Members does not increase capacity – ten people sharing one slot still process one Task at a time.

Model usage

Model spend is separate from slots and is either prepaid as credits or billed directly by your own provider under BYOK. Coroid does not add any surcharge to provider token rates.

Running out of credits halts new model calls. It does not affect your slots, Projects or data.

See Model credits.

Rate limits

API and MCP access are rate limited per organisation. These limits are generous for interactive use; if you are driving Coroid programmatically at scale, expect to handle 429 responses with backoff.

What occurs at a limit

No limit results in data deletion. Reaching a limit stops new work from starting, and the interface informs you which limit was hit. Existing Projects, Tasks, history and settings remain untouched.