Why is CRHQ.ai making $18.5k per month?

Is CRHQ.ai really making $18.5k per month?

CRHQ.ai making roughly $18.5k per month looks credible today, although we can verify the direction of the revenue more confidently than the exact latest figure.

FounderBase currently lists CRHQ at $14,331 in monthly recurring revenue. A later public revenue breakdown from founder Andrej Šimunaj put CRHQ at around $16,500 for July. Getting from there to $18,500 would mean another roughly 12% increase.

That kind of move does not require a sudden viral breakout. CRHQ currently charges $299 per month for its standard satellite, with larger servers at $349 and $399. At $299, another $2,000 of MRR is roughly seven additional satellite subscriptions.

The three revenue points fit together reasonably well. They look like a business adding a handful of substantial subscriptions at a time rather than hundreds of low-priced customers.

We should still distinguish between a founder-reported run rate and audited company revenue. CRHQ does not publish Stripe data, customer counts or monthly accounts that would let us independently reconcile the $18.5k figure.

For now, the stronger claim is that CRHQ has clearly reached the mid-five figures in monthly recurring revenue and appears to still be moving upward.

CRHQ revenue reference	Approximate monthly revenue	$299 satellite equivalents
FounderBase snapshot	$14,331	48
Later founder revenue breakdown	~$16,500	55
Current reported run rate	~$18,500	62

What does CRHQ.ai actually sell for $299 a month?

CRHQ.ai currently sells a permanent workplace for AI agents: a dedicated server with memory, tools, files, browser access, scheduling and agent management already wired together.

Each CRHQ “satellite” runs on its own virtual private server. According to CRHQ’s current technical documentation, the environment includes PostgreSQL, a file system, persistent agent sessions, a browser, encrypted credentials and scheduled jobs.

CRHQ has also started exposing more of that infrastructure directly to developers. Its newer “Build on CRHQ” documentation lets another product use a CRHQ instance as its agent runtime. Sessions can persist, requests are queued safely, agents keep their configured skills, and interrupted sessions can resume after a server restart.

That makes the product easier to understand. A customer could assemble many of these components independently with a VPS, Claude Code or Codex, PostgreSQL, cron, browser automation and some custom glue code. CRHQ packages the stack and maintains it.

The customer still chooses and pays for the underlying AI. CRHQ currently supports external model access rather than bundling unlimited Claude or OpenAI usage into the $299 fee.

So when a customer pays CRHQ every month, most of that payment is for the environment surrounding the model: persistence, deployment, orchestration and maintenance.

How can CRHQ reach $18.5k without having thousands of customers?

CRHQ can reach $18.5k with surprisingly few deployments because its starting price is already $299 per month.

At that price, $18,500 corresponds to roughly 62 standard satellites. If some customers use the $349 or $399 tiers, the equivalent number drops into the 40s or 50s.

The customer count can be lower still because CRHQ bills per satellite rather than strictly per company. One organization may run several environments.

CRHQ’s own product example currently shows separate research, development, content and operations satellites managed from one Hub. A company using four standard satellites would generate almost $1,200 a month.

Agency deployments push the same logic further. CRHQ now supports custom wildcard domains, so an agency can provision separate environments such as client1.agency.com and client2.agency.com while managing them centrally.

Once you look at the pricing, the revenue stops looking strange. A $19 SaaS product would need close to 1,000 subscriptions to reach $18.5k. CRHQ needs dozens of active environments.

Monthly price	Subscriptions needed for $18.5k MRR
$19	974
$99	187
$299	62
$349	53
$399	47

Why are people willing to pay CRHQ $299 when they can run Claude Code on a cheap server?

Some CRHQ customers are paying $299 because maintaining an agent stack themselves would save money but create another system they have to operate.

CRHQ openly acknowledges the trade-off. Its current Claude Code comparison says a technical user can run Claude Code on a cheap VPS and get a useful always-on coding setup.

Getting closer to CRHQ means adding quite a lot around it: a web interface, persistent projects, memory, credential storage, scheduling, browser sessions, backups, monitoring, audit history, reusable skills and multi-server administration.

A competent developer can build those pieces. The more relevant question for CRHQ’s target customer is whether spending engineering time on them is worth saving roughly $300 a month.

Founder Andrej Šimunaj recently made essentially the same argument about SaaS more broadly. He wrote that even though his agents can recreate many small SaaS tools, he still pays for inexpensive products such as Calendly because rebuilding something that saves $29 a month is a poor use of the system.

CRHQ applies that logic one level higher. A founder running several revenue-generating agents may happily spend $299 to avoid maintaining the infrastructure those agents depend on.

For a hobbyist, the price is difficult to justify. For someone already spending thousands of dollars on employees, contractors, APIs and software, $299 lands in a very different part of the budget.

Did CRHQ start with a real use case, or was it built because AI agents became fashionable?

CRHQ appears to have grown out of Andrej Šimunaj’s own operating setup, and that origin probably helped it find paying users quickly.

Before CRHQ became a commercial product, Šimunaj was publicly documenting increasingly large projects built with autonomous agents. One example was a complete travel-guide website that he said his agents assembled in four days, including software, accommodation data and content.

A few weeks later, he described turning the underlying setup into CRHQ after people reacted strongly to those experiments. During the private-alpha phase, access was still being given manually through a waitlist.

His more recent posts show the same workflow continuing. He has described running roughly 10 to 15 agents and letting them work while he is away. In another recent example, he said an API product was built while he was on holiday and was later launched with paying usage.

There is an obvious limitation here: these are the founder’s own accounts of what his system produced. They are more useful as evidence of product origin and sales positioning than as independent proof of agent productivity.

Still, they explain why the first buyers existed. CRHQ was being demonstrated through the exact workflow it wanted customers to adopt.

Is Andrej Šimunaj’s existing business helping CRHQ make money?

Yes. CRHQ benefits from a founder who already had products, customers and a software business before launching the agent platform.

CRHQ is operated by Zero Point Studio, Šimunaj’s Croatian software company. Croatian company-data services report that Zero Point Studio generated about €143,000 of revenue in 2024 and around €280,000 in 2025.

Those figures cover the wider company rather than CRHQ specifically, so we should not mix them into CRHQ’s MRR. They do show that the business behind CRHQ did not begin from zero.

Zero Point Studio also operates products including TranscriptAPI, Recapio and Cascady. Its current website says TranscriptAPI processes more than 15 million transcripts per month, while Recapio has more than 55,000 active users.

That existing product portfolio gives CRHQ two advantages. Šimunaj already understands how to launch and operate internet products, and those products provide live environments where he can test the agent workflows he later sells through CRHQ.

This helps explain why CRHQ could reach five-figure MRR quickly despite being young. The software is new; the commercial starting point around it is much more mature.

Is founder-led distribution doing most of the work for CRHQ?

Founder-led distribution has clearly driven CRHQ’s early sales, and at its current size that can be enough.

Šimunaj has spent months posting detailed examples of his agent workflows on LinkedIn and X, replying to other developers, showing products made with agents and inviting interested people into CRHQ.

During private alpha, he explicitly told people in his network to message him for priority access. In another public discussion about an AI workshop, he directly suggested CRHQ to the organizer and offered to jump on a call.

This is a concentrated acquisition channel, but CRHQ does not need a massive audience because its contract value is high.

Suppose a technical post reaches 5,000 relevant founders, agencies and developers. Converting just 0.2% of those people into $299 customers would mean ten customers and almost $3,000 of additional MRR.

That rough example is illustrative rather than a claim about CRHQ’s actual conversion rate. It shows why a founder with a few thousand highly relevant followers can build meaningful SaaS revenue without becoming a mass-market influencer.

The harder test comes later. Founder-led selling can get CRHQ into the tens of thousands of dollars per month. Reaching much further will require referrals, search traffic, agency expansion or another acquisition engine that does not depend so heavily on Šimunaj personally.

Why does CRHQ’s $299 pricing work so well financially?

CRHQ’s current pricing leaves a lot of room between what a satellite costs to operate and what the customer pays for it.

The standard $299 plan includes a 4-CPU, 8-GB RAM VPS. CRHQ’s technical documentation says these instances run on Hetzner Cloud, where comparable raw servers cost far less than $299 a month.

CRHQ naturally has costs beyond the VPS: development, backups, monitoring, databases, support, security work and payment processing all have to be paid for.

The expensive variable in many AI applications, model inference, sits largely outside CRHQ’s own bill. Customers connect their own Claude subscription, Codex subscription or API credentials and pay the AI provider directly.

FounderBase currently reports an 89% net margin for CRHQ at its $14,331 MRR snapshot. If that figure is measured consistently, it implies roughly $12,750 of monthly profit and around $1,580 of total monthly costs at that point.

We would not assume an 89% net margin survives once CRHQ hires more people, provides heavier support and serves larger companies. The basic spread between the platform price and the direct infrastructure cost is the more durable part.

FounderBase CRHQ snapshot	Approximate figure
Monthly revenue	$14,331
Reported net margin	89%
Implied monthly net profit	~$12,755
Implied monthly costs	~$1,576

Does CRHQ make more money when customers use more AI?

CRHQ can benefit from heavier customer usage without directly taking on most of the extra model cost, which gives the business unusually clean economics for an AI product.

A customer pays CRHQ for the satellite and pays Anthropic, OpenAI or another model provider separately. More Claude or GPT usage therefore does not automatically create an equally large expense on CRHQ’s own income statement.

This became especially visible when AI-provider billing rules changed. CRHQ could publish guidance telling customers how to adjust providers or authentication rather than having to absorb a sudden model-cost increase into an all-inclusive subscription.

The setup also makes CRHQ less dependent on one model company. Its current materials reference Anthropic, OpenAI, Bedrock and Vertex options, while the product can evolve as the best model changes.

Customers do have to tolerate multiple bills. A heavy agent user could pay $299 to CRHQ and then spend considerably more on AI access.

That probably narrows the market, but it also selects for customers doing valuable enough work to care more about the result than about keeping their total software bill under $50.

Could agencies become CRHQ’s best customers?

Agencies could become especially valuable CRHQ customers because one agency can turn a single relationship into several paid satellites.

CRHQ’s newer custom-domain system is clearly designed with this behavior in mind. An agency can connect its own domain and give individual client satellites branded addresses rather than sending every customer to a CRHQ-branded URL.

The central Hub can then manage those separate environments as a fleet.

This changes the economics for the buyer. An agency using CRHQ only for its internal work has to justify a $299 monthly productivity expense. An agency packaging a satellite into a client service can recover that cost inside a much larger monthly retainer.

Five client satellites at the standard price are worth $1,495 of MRR to CRHQ from one agency relationship. Ten are worth $2,990.

CRHQ does not currently publish the split between single-satellite customers and multi-satellite organizations, so we cannot say how much of today’s revenue already comes from this pattern.

The product is increasingly built for it, though. White-label domains, fleet management and dedicated isolation make considerably more sense for agencies and client-service businesses than for someone who simply wants one personal AI assistant.

What can CRHQ do today that cheaper agent tools do not make easy?

CRHQ currently combines managed infrastructure, business-oriented agent tooling and multi-environment control in one place, which is where its premium price still has some justification.

OpenClaw is an obvious comparison. CRHQ’s latest comparison page says OpenClaw has roughly 368,000 GitHub stars, more than 1,200 contributors and a marketplace containing thousands of skills. It can be self-hosted for free, while its hosted cloud option is advertised at $59 per month.

CRHQ starts at $299.

Paying five times as much only makes sense if the customer cares about what CRHQ adds: managed patches and backups, an operations interface, dedicated VPS isolation, a credential vault, curated business skills, named agent roles, fleet management and agency white-labeling.

The gap is moving. CRHQ recently added an API-oriented runtime that lets applications build directly on its sessions, agents and skills. Meanwhile, OpenClaw and large AI vendors keep adding capabilities of their own.

CRHQ’s advantage today is a bundle of operational conveniences rather than some unique form of artificial intelligence.

That can support a good software business. Plenty of valuable infrastructure companies sell packaging, reliability and easier operations around technologies customers could theoretically assemble themselves.

Could Claude, Codex, Grok or OpenClaw make CRHQ unnecessary?

Yes, platform convergence is probably the biggest threat to CRHQ’s current $299 price.

Claude Code and Codex keep becoming more autonomous. OpenClaw offers always-on agents with an enormous open-source ecosystem. CRHQ recently added a comparison with Grok Bot because xAI is pushing persistent agent behavior into its own stack as well.

CRHQ’s own comparison pages make the pressure unusually visible. The company now has dedicated pages explaining why someone should choose CRHQ over Claude Code, Codex, OpenClaw and Grok Bot.

That is useful marketing, but it also tells us exactly where the product can get squeezed.

If frontier-model vendors eventually bundle durable memory, scheduling, browsers, credential management, multi-agent workflows, persistent hosting and team administration into subscriptions customers already pay for, a separate $299 layer becomes harder to sell.

CRHQ therefore needs to move faster than those generic platforms in areas they care less about. Agency fleets, white-label environments, self-contained business runtimes and highly specific operational skills are plausible places to do it.

As of now, that gap still exists. We would be much less confident that the same gap will automatically exist several years from now.

Are CRHQ customers likely to keep paying every month?

CRHQ has good ingredients for retention, although we still lack the churn data needed to say that retention is already proven.

A satellite becomes more useful as a customer fills it with projects, credentials, scheduled jobs, custom skills, browser sessions, databases and agent memory.

CRHQ’s newer developer runtime deepens that attachment. A company can now build product functionality around persistent CRHQ sessions, agent configurations and queues rather than using the service only as a chat interface.

Once those workflows are running every day, cancelling the subscription means more than losing access to a generic AI chat window. Someone has to move or rebuild the operating environment.

The lock-in remains manageable. CRHQ says customers own their workflows and data, and technically capable teams can recreate much of the stack elsewhere.

That combination should help retention: there is enough operational friction to make switching annoying, while the product cannot rely on trapping the customer.

What we cannot see today is the most important number: how many of the first CRHQ customers are still paying three, six or twelve months later. Until that cohort data exists, strong retention remains a reasonable hypothesis rather than a demonstrated fact.

Is CRHQ’s growth coming from AI hype or actual customer value?

AI hype clearly helped CRHQ get attention, but the pricing makes it difficult for hype alone to support the business for very long.

A curious user can subscribe to a $20 AI app, play with it twice and forget to cancel. A $299 monthly satellite creates a much higher bar.

At that price, customers eventually have to connect CRHQ to something economically useful: software development, research, SEO, reporting, content operations, monitoring or client delivery.

CRHQ’s current use-case library reflects that shift. The company now presents workflows such as weekly competitor monitoring, technical SEO audits, Stripe and analytics reporting, content scheduling and database-backed business intelligence.

Those examples are more commercially useful than broad promises about having a “team of AI employees.” They also make the buying decision easier to evaluate. If a recurring competitor report saves five hours of expensive work every week, $299 is easy to defend. If the satellite mostly produces interesting conversations, it is expensive.

The next stage of CRHQ’s growth will tell us which version dominates. Five-figure MRR proves that a group of customers already sees enough value to pay. It does not yet prove that the same behavior will spread to a much larger market.

Can CRHQ.ai keep growing beyond $18.5k per month?

CRHQ can realistically grow well beyond $18.5k per month because the next revenue milestones still require a manageable number of additional deployments.

At $299 per satellite, moving from $18,500 to $30,000 means adding roughly 39 standard-equivalent satellites. Reaching $50,000 from the current level takes about 105 more.

An agency that eventually operates ten client satellites could account for almost 10% of that entire jump to $50,000 by itself.

CRHQ also has more product surface today than it had during its private-alpha period. The platform now has detailed documentation, reusable skills, recipes, white-label domains, fleet administration and an API runtime that other products can build on.

Those additions give CRHQ more ways to become embedded in a customer’s work.

Competition is moving just as quickly. OpenClaw, Claude, Codex, Grok and other agent platforms are steadily making persistent autonomous work easier and cheaper.

Near-term growth should depend less on finding thousands of users and more on three things: keeping the early customers, getting existing organizations to add satellites and convincing agencies or software companies to build recurring work around the platform.

That is a credible route from $18.5k to considerably more. It is also a much harder test than reaching the first few dozen deployments.

So why is CRHQ.ai making $18.5k per month?

CRHQ.ai is making roughly $18.5k per month because a few dozen serious deployments can already produce that revenue, and CRHQ has built a product that some founders, agencies and technical teams consider worth several hundred dollars every month.

The pricing does a lot of the work. A standard satellite costs $299, so the current run rate corresponds to only about 62 standard-equivalent subscriptions. Larger plans and organizations running several satellites can reduce the number of actual customers substantially.

The economics are equally important. Customers pay their own AI-provider bills, while CRHQ supplies the server and orchestration layer. FounderBase’s reported 89% net margin may fall as the company grows, but it shows how profitable this setup can be at small scale.

Then there is distribution. Andrej Šimunaj spent months publicly showing the same autonomous workflows CRHQ eventually packaged. Zero Point Studio already had operating products and paying customers, so CRHQ launched from inside an existing software business rather than from a blank page.

Lately, CRHQ has also been moving deeper into infrastructure. White-label client environments, fleet management and the newer agent-runtime API give customers reasons to build more of their operations around the platform.

The $18.5k figure looks mostly credible rather than mysterious. We are looking at a high-priced young SaaS product that only needs dozens of active environments to reach five-figure MRR.

The harder question starts now. CRHQ has shown that people will pay $299 for managed agent infrastructure while building the same stack yourself is still annoying. Its next challenge is keeping that convenience valuable as Claude, Codex, Grok, OpenClaw and the rest of the agent market keep closing the gap.