Pricing

Simple pricing.
Scale by the agent.

$60/mo gets you Hydrogen and two specialists. Add any of the other 22 for $25/mo each. Compute and AI usage are billed separately — your own instance, your own key, no bundling and no markup.

Subscription

$60/ month

Hydrogen + 2 specialists of your choice

Agents: Add any of the other 22 for $25/mo · change anytime
API key: Yours — Anthropic, OpenAI, or Google, no markup
Runs on: Your own dedicated AWS server — from ~$30/mo, passed through at cost
Annual: 2 months free on base + agents

Build your team ↓

AI usage always runs on your own key for paid plans — billed by your provider, never marked up by us. Compute is AWS's own rate, passed through — no markup.

Build your team. Estimate your cost.

Pick your specialists and server size. This estimate includes subscription and compute; your AI provider bills usage separately.

Extra specialists

$25 × 3 = $75/mo

0510152022

Example configuration: 5 specialists on a Standard server. Adjust it to your workload.

Compute — your server size

Paid to AWS at cost, no markup

Or pick your team by name

Estimated monthly cost

$195/ mo

~$2,340/year billed monthly

Base — Hydrogen + 2 included$60/mo

3 extra specialists × $25/mo$75/mo

Compute — Standard (t3.large)~$60/mo

Billed separately

AI usage — your own keyvaries with volume

Compute is AWS's rate, passed through at cost and billed monthly. On top sits your provider's token bill — Hydrogen's QA pass is part of that usage; every output is generated and reviewed.

Request access

Compute

Your team runs on your own machine.

Every paying customer gets a dedicated server on AWS. You pay AWS's own rate for it — passed through, no markup, just like your API key. Change size anytime, one click, prorated.

Light

Solo founders · light workloads

~$30/ month, paid to AWS

AWS t3.medium

Your team handles a few briefs a week comfortably. The right start if you're one person testing the water.

Standard

Most teams

~$60/ month, paid to AWS

AWS t3.large

Handles daily briefs with several specialists working together. Where most agencies and small teams land.

Performance

Heavy, continuous work

~$140/ month, paid to AWS

AWS m5.xlarge

Your full roster runs in parallel without queueing. For teams that keep the workspace busy all day.

What if I pick too small?

A smaller instance means your team works through briefs more slowly. It never changes the quality bar — Hydrogen reviews everything the same way on every tier.

Upgrades

One click, effective immediately, prorated. Prices are AWS's published rates and can shift slightly — we never add a cent on top.

Your own instance.
Nobody else's data.

“Can I put client data in here?” — yes. Every paid workspace runs on dedicated, isolated infrastructure: your environment, your data, your key. Nothing shared, nothing pooled.

Dedicated instance per customer

Your work never shares a machine with another customer's.

Your key, your provider terms

AI traffic goes to the provider you chose, under your account.

A reality check

A junior content hire: $4,500+/month, one draft at a time.
Nitrogen: $25/month — every draft reviewed by Hydrogen first.

Junior content marketer, fully loaded US cost · Nitrogen added to any Space Office subscription

Illustrative review example

What a review should change.

This example shows the kind of feedback a review can provide. It is not a recorded customer task or a measured success rate. You still check and approve the final work.

See the task flow Browse specialists

1. Your brief

Write a launch announcement for a $19 desk organizer. Use a calm tone. Do not make shipping or sustainability claims; neither is confirmed.

2. Draft needing review

The world's most sustainable desk organizer is here. Just $19, with guaranteed next-day delivery!

3. Review notes

Remove the unsupported sustainability superlative and delivery promise. Keep the confirmed $19 price and use the requested calm tone.

4. Revised draft

Meet our new $19 desk organizer. Give your everyday essentials a place of their own. Explore the organizer and see whether it fits your workspace.

Common pricing questions

Things people ask.

$60/month for Hydrogen — your AI project manager — plus two specialists of your choice. Add any of the other 22 for $25/month each. On top of that you pay AWS directly for your server (from about $30/month) and your AI provider directly for token usage. We never mark either up.

It was a flat rate that bundled compute into the subscription, which meant light users subsidised heavy ones. Splitting it out means the base drops to $60, you only pay for the server size you actually need, and the AWS and provider costs reach you at cost instead of through us.

Every workspace runs on its own dedicated AWS instance rather than a shared pool. Passing AWS's published rate straight through means you scale the machine to your workload — and you can verify what you're paying against Amazon's own pricing.

Your team works through briefs more slowly. It never changes the quality bar — Hydrogen reviews everything the same way on every tier. Upgrading is one click, effective immediately, and prorated.

Yes, for paid plans. You connect your Anthropic, OpenAI, Google account and AI traffic runs under it, so you pay the provider directly with zero markup from us. Hydrogen's QA pass is part of that usage — every output is generated and reviewed. A managed option is on the roadmap.

Yes. Monthly plans cancel at the end of the current month with no fee. Annual plans have a prorated refund window for 30 days; after that they run to term. Cancellation is handled from the dashboard, not a retention call.

Yes. Paid Space Office workspaces run on a dedicated AWS instance. Compute is billed separately at AWS cost, and you can resize as workload changes.

Ready to build your team?

Request access →

Simple pricing.Scale by the agent.