Digital Teammates, No Surprises.
One‑time setup and transparent monthly costs. No per‑user fees.
Pricing overview
No per‑user costs.
One appliance powers up to 25 AI coworkers, each with up to 8 apps — up to 200 apps on a single box.
Unlimited app usage.
10% annually per paid app (billed monthly).
Token usage is included for most apps. For high-volume scenarios, we may pass through usage at cost (with notice) so your spend stays predictable.
Calculator
Minimum 1
Setup includes 1 app(s); others billed.
How pricing works
Five pieces, zero surprises
Every customer starts with the same building blocks. Pick how many AI coworkers and workflow apps you want, and the math stays simple as you grow.
Includes your first AI coworker and its first workflow app. We install, secure, and test inside your Microsoft 365 tenant.
Why it matters
Covers all the heavy lifting upfront so you start with a working teammate on day one.
A managed on-premise appliance that runs Agents locally. One box powers up to 50 AI coworkers and 400 workflow apps without opening new ports or exposing data.
Why it matters
Keeps everything privacy-first — local inference ready, no external portals, and still a flat infrastructure cost with zero surprise GPU bills.
Each AI coworker is a Microsoft 365 user scoped to a specific domain (think HR, finance, operations) with its own email address, Teams presence, calendar, and file space.
Why it matters
IT provisions and governs agents just like any employee account, so security stays simple while you spin up domain experts on demand.
Every new workflow app, integration, or automation is a one-time build fee after your first included workflow app, with unlimited usage once live.
Why it matters
Add automations over time without increasing your monthly base or worrying about per-run charges.
Every workflow app has a one-time build of $5,000 and a flat upkeep of $500 (10%) billed monthly at $42. Covers updates, prompt tuning, and monitoring.
Why it matters
You keep predictable care-and-feeding costs without negotiating support each time you add a workflow app.
Unlimited usage is included. Costs only move when you add agents or apps, so finance can model spend without guesswork.
Launch with one coworker, then layer on more workflow apps as you prove value. Each addition is a clear line item.
Everything runs in your tenant on hardware you control, which keeps data residency, audit, and compliance boxes checked.
How the numbers roll up
Two quick examples customers ask for most
Use these as a gut-check, then tweak the calculator with your own agent and app counts.
New customer with one agent and a single workflow app (included in setup).
- First-month total
- $11,000
- One-time at kickoff
- $10,000
- Monthly recurring
- $1,000
- Monthly maintenance
- $0
- Year 1 total
- $22,000
- 1Setup covers your first agent and workflow app, so there are no app fees or maintenance yet.
- 2Monthly spend is hardware plus one agent subscription.
New customer launching three agents across five workflow apps.
- First-month total
- $31,667
- One-time at kickoff
- $30,000
- Monthly recurring
- $1,500
- Monthly maintenance
- $167
- Year 1 total
- $50,000
- 1Four additional workflow apps are billed once at kickoff and carry monthly maintenance.
- 2Monthly recurring scales with agent instances; maintenance tracks the workflow apps you add.
What is an Agent?
Familiar Tools. Real Work.
Each agent is a Microsoft 365 user (e.g., agentaech@yourcompany.com). Email or message in Teams—no portals, no new UI.
Receives work by email—send what you’d email a coworker (receipts, invoices, resumes, contracts) and it converts, organizes, and routes it to the right systems.
Receives work in Teams—chat or @mention to assign tasks and get results back in‑channel with updates and files.
Schedules its own tasks, runs, and reports on a cadence — no nudges needed.
Stores outputs, writes docs, and organizes artifacts in its space, honoring your permissions and retention.
You talk to it like any colleague — email or Teams. That’s the interface.
Provisioned as an M365 user you control; respects your MFA, DLP, groups, and audit trails.
Frequently asked questions
- What is Agent Aech?
- A digital teammate that communicates over email and Microsoft Teams to deliver real work. No portals, no new UI—just the tools your team already uses every day.
- Where does it run?
- On your network, on the Agent Aech hardware appliance (Mac Studio Ultra, 96GB-512GB RAM). It is securely connected and fully managed, so you get performance without operational overhead.
- Does our data leave our network?
- Your data stays on your network. Only the minimal context required to generate a response is sent to the LLM via secure enterprise APIs with no training and no data retention. Nothing is persisted by vendors.
- How is access controlled?
- Each agent is a Microsoft 365 user you manage (e.g., agentaech@yourcompany.com). Grant access by adding it to groups, mailboxes, and folders—just like any employee. It honors your MFA, DLP, retention, and group policies.
- What can it do out of the box?
- Read and write email, join Teams conversations, schedule and send reports on a daily/weekly cadence, and organize files in OneDrive/SharePoint. Custom apps automate your specific workflows.
- Do we need to install new software?
- No. You talk to the agent via email and Teams. Admins manage it like any Microsoft 365 user.
- How is pricing structured?
- No per-user fees. Pricing includes a flat hardware subscription and a per-agent monthly fee. Apps are unlimited use; you pay a one-time fee per app plus maintenance (10% annually per paid app).
- How long does setup take?
- We provision the appliance and your first agent, then connect the right permissions. Most teams are live in days, not months.
- How do token costs work?
- Token usage is included for most apps. If a specific workload generates exceptionally high token volume, we may pass through usage at cost (with notice) so your spend stays predictable. We’ll help you tune prompts and batching to keep usage efficient.
- Why a powerful appliance, and what’s the roadmap?
- We’ve spec’d the Agent Aech appliance (Mac Studio Ultra, 96GB+ RAM) to run more and more AI workloads locally over time. Today, only minimal context leaves your network to LLMs via secure, no‑retention enterprise APIs. Our roadmap pushes toward on‑appliance inference so you gain lower latency and stronger privacy. The vision: 100% of AI workloads run locally, 100% of your data stays private.