Plans & Limits

All plans include unlimited team seats at no extra cost.

New accounts start with a 14-day free trial with full Team-tier access. No credit card required.

Pricing

	Free	Team	Business	Enterprise
Annual	€0	€499/mo	€1,499/mo	Custom
Monthly	€0	€599/mo	€1,799/mo	Custom

Annual pricing is the base rate. Monthly billing is 20% higher, no lock-in.

Plan comparison

Feature	Free	Team	Business	Enterprise
Team seats	Unlimited	Unlimited	Unlimited	Unlimited
Active flows	3	25	Unlimited	Unlimited
Messages/mo	5,000	50,000	250,000	Custom
AI actions/mo	50	500	2,000	Custom
AI action packs	—	—	€99 / 500	Included

Standard connectors	3	All	All	All
Custom connectors (prod)	1	5	Unlimited	Unlimited
Custom connector gen cost	5 AI actions	5 AI actions	3 AI actions	1 AI action
Connector sharing (org-wide)	—	—	Yes	Yes
Private connector registry	—	—	—	Yes

Self-healing	Included (fixed 95%)	Included (configurable)	Included (configurable)	Included (configurable)
Environments	1	2	5	Unlimited
API publishing	—	Yes	Yes	Yes

Governance (PII, residency)	—	—	Yes	Yes + custom policies
SSO / SAML	—	—	Yes	Yes
Audit log exports	—	—	Yes	Yes

Support	Community	Email 24h	Slack 4h	Dedicated 1h + CSM
SLA	—	99.5%	99.9%	99.95%+

What counts as an AI action?

An AI action is one AI generation, iteration, or connector build performed via the CLI, MCP, or web chat. Examples: generating a flow from a prompt, iterating on a mapping, building a custom connector.

Deploying and running flows does not consume AI actions. Once a flow is generated, it runs on the deterministic runtime at no AI cost.

Custom connector generation costs AI actions per build: 5 on Free and Team, 3 on Business, 1 on Enterprise.

Connectors

Paid plans (Team, Business, Enterprise) include access to all standard connectors with no per-connector fees. The Free tier is limited to 3 standard connectors and 1 custom connector in production.

Business and Enterprise plans support org-wide connector sharing, so a connector configured by one team member is available to everyone. Enterprise adds a private connector registry for governed, versioned connector distribution.

Rate limits

Resource	Free	Team	Business	Enterprise
Webhook ingestion	10 req/s	100 req/s	500 req/s	Custom
API requests	60 req/min	600 req/min	3,000 req/min	Custom
MCP tool calls	10 req/s	50 req/s	200 req/s	Custom
AI generations	5 req/min	30 req/min	60 req/min	Custom
Batch concurrency	2	10	25	50
Published API calls	—	1,000 req/min	5,000 req/min	Custom

Payload limits

Limit	Value
Max payload size	10 MB (configurable per flow via `limits.max_payload_size`)
Max array size	Configurable per flow via `limits.max_array_size`
Auto-chunking	Configurable per flow via `limits.chunk_size`
Max wait duration	7 days
Max pagination pages	100 (per poll cycle)
Max array map nesting	5 levels
Max aggregate buffer	10,000 messages (configurable per flow)

Duration limits

Duration field	Max value
`wait` step	7 days
`dedup.window`	No hard limit (recommended: ≤ 30 days)
`ordering.lock_ttl`	No hard limit (default: 60s)
`aggregate.window`	No hard limit (recommended: ≤ 1 hour)

Connector rate limits

Per-connector instance rate limiting is configurable:

# Set via API or MCP when creating a connector instance
rate_limit_config:
  requestsPerSecond: 2
  requestsPerMinute: 120
  dailyQuota: 10000
  burst: 5
  onLimit: queue    # queue | reject | backoff

Behavior	Description
`queue`	Buffer requests and send when capacity is available
`reject`	Return an error immediately
`backoff`	Wait and retry with exponential backoff