Pricing
Free for one workflow. Team from $29 per user, per month.
Connect real work, record its cost and acceptance baseline, and choose a paid plan only when your team needs more workflows or stronger controls.
$0
One live workflow and one shared board.
Agents are not billable seats; they are participants.
One live workflow
Use one real operating path.
The board
One source of truth for people and agents.
Your agents as assignees
Work goes to whoever does it best.
Goals connected
Keep the goal attached to the work.
The baseline measurement
Record cost, time, outcome, and acceptance.
Workspace memory with cited claims
Give every agent context that links back to the work it came from.
$29
per human user, per month
More workflows, boards, planning, and reporting.
Run more operating paths and compare their results.
Unlimited workflows and boards
Every operating path shares one work model.
Planning and reporting across goals
Progress rolls up without a program manager.
Acceleration on every workflow
Compare runs, evaluate candidates, and control releases.
Workspace memory with cited claims
Share decisions and rules across the team with their evidence attached.
$99
per human user + per governed workflow
Per-workflow price announced at launch
Everything in Team, plus policy, approval, and evidence.
Keep the decision boundary and proof with governed work.
Everything in Team
So governance does not remove the operating gains.
The policy engine
Pin the active policy boundary to each governed run.
Approval gates
Hold supported consequential actions for a named person.
The tamper-evident evidence record
Link every decision in one receipt chain.
Governed sandbox hours
Run coding agents on a machine your policy defines, with hours included per seat.
Organization-wide memory, approval gate, retention policy
Promote shared rules once, review what agents can use, and control how long their sources remain available.
Talk to us
SSO, deployment options, custom policy models.
The governance review and the ROI worksheet run with your numbers.
Everything in Governed
Keep organization memory, its approval gate, and its retention policy.
SSO
So access follows the identity system you already trust.
Deployment options
So data and runtime boundaries fit your operating model.
Custom policy models
So your controls become decisions inside the workflow.
The governance review and ROI worksheet
So the rollout starts with your risks and your numbers.
Bring your own compute
Sandboxes in your own cloud account. On the roadmap.
Memory scoped per team, export, kill switch controls
Keep team-private claims with current members, export memory, or stop serving it.
Set people, agents, workflows, and runs. The estimate uses the active plan terms and Meridian's run curve.
Meridian Freight · Ship the governed customer import
Lands in
Monthly estimate
Modeled from your inputs on the active plan terms. The team, default usage, and run curve are fictional examples, not measured customer costs or savings. Plan prices are separate from these example inputs.
No. Human users are seats. Agents are workflow participants.
No. Keep them and connect one workflow first.
Start with one workflow that already has a clear request and accepted result. Security-led teams can start with one approval path and its evidence record.
Your workflows, context, and evidence records are yours.
Security and data handlingUse models available through your connected providers and supported agent runtimes. A workflow can choose a different model or worker for each step.
Each human user carries an allowance of sandbox hours, and hours past the allowance are metered on two things: the CPU a sandbox actually uses and the memory it holds while it is provisioned, so a session waiting on a model response is not billed as though it were computing. An idle sandbox snapshots and stops, and its snapshot storage passes through at cost. Rates are announced with the rest of the plan terms.
Aestus starts with real work and a named acceptance decision, not a demo prompt. Gartner (2025) expects more than 40% of agentic projects to be canceled because of cost, unclear value, or weak risk controls.