What would stop you from trying ComfyOnline (https://www.comfyonline.app) this month? Be specific about how you handle this today. Also say the maximum you would pay per month in USD (0 if nothing).

Page reviewed

Run ComfyUI workflows online and deploy APIs with one click - ComfyOnline

https://www.comfyonline.app/ (opens in a new tab)
Saved screenshot of Run ComfyUI workflows online and deploy APIs with one click - ComfyOnline
Screenshot saved with this question. The live page may have changed.Open full image ↗

10 fictional AI personas. Exploratory perspectives, not a survey of real customers. Methodology & limitations

Simulated interest

45/100

AI synthesis, not measured demand

Suggested price range

$0–150

Median $50 · simulated willingness to pay

Response sentiment

Positive 0%Neutral 100%Negative 0%
The takeaway

What this panel suggests

The responses are broadly neutral: people see the value of running ComfyUI online and generating APIs, but they are not ready to try ComfyOnline this month because they cannot estimate the cost or the migration risk. The strongest blocker is pricing clarity. Several respondents said the page’s “pay only for runtime” message sounds good, but they still need actual per-hour or per-tier numbers to understand what they would pay in practice. That concern is amplified by the presence of H100/A100/4090 on the page without enough detail to model monthly spend. A second concern is workflow portability: users want confidence that their existing custom nodes, graphs, and granular control will work as expected. Some respondents also said they already have adequate local or shared-machine setups, including home GPUs, Docker, or cloud credits, so switching has to be clearly better. Stated maximum monthly willingness to pay spans from $0 to $150, with a median of $50. This is a stated ceiling from the synthetic answers, not proof of demand or a recommended price.

A hypothesis to test

Pricing & positioning

Lead with a transparent runtime cost calculator and publish example hourly/monthly costs for the main GPU tiers. Add a compatibility checklist or import test for custom nodes and workflows before pushing users toward a paid plan.

The main friction is not rejection of the product idea; it is uncertainty about what a month will cost and whether existing workflows will перенос cleanly. A capped first month, free trial, or spend limit would likely reduce perceived risk.

Recurring themes

Pricing transparency and runtime cost uncertainty
“the pricing isn't super clear from the page – 'pay only for runtime' sounds good, but I'd need to understand the per-hour costs”Read Jasmine Heidenreich’s response →
Workflow compatibility and custom node migration risk
“ensuring compatibility with my existing custom nodes and maintaining that level of granular control”Read Lewis Waelchi’s response →
Existing infrastructure and sunk costs
“My current setup with a home GPU and Docker is quite efficient”Read Lewis Waelchi’s response →

Objections to investigate

  • Unclear pricing structure
  • Custom node and workflow compatibility risks
  • Sufficient existing local/cloud setup
  • Lack of specific GPU tier details

Use these findings to plan customer interviews or experiments. How to interpret AI research

What the panel said

Click a name to open that persona’s full profile.

Lewis Waelchineutral

My current setup with a home GPU and Docker is quite efficient for my specific workflow needs, especially since I meticulously track node versions and configurations. The primary hesitation for me would be ensuring compatibility with my existing custom nodes and maintaining that level of granular control I've built into my processes. I'd need to see clear evidence that ComfyOnline doesn't introduce limitations or complexity that outweighs the benefits of not managing my own hardware. Given my focus on efficiency and reproducibility, I'm not looking to add another variable unless it offers significant advantages. I would consider paying up to $50 per month if it demonstrably streamlines deployment or offers access to hardware I can't easily replicate.

Simulated interest: 40/100

Jasmine Heidenreichneutral

Honestly, the biggest thing holding me back is my current setup with RunComfy. It's working, and I've ironed out most of the kinks for my weekly sprints. I'm all about optimizing my pipeline, so before I switch, I'd need to see how ComfyOnline integrates with my existing Figma and Webflow workflow, especially for demos. Also, the pricing isn't super clear from the page – 'pay only for runtime' sounds good, but I'd need to understand the per-hour costs for specific GPUs like the A100 to gauge if it's more efficient than my current setup. I'm willing to pay up to $150 per month in USD, but that's dependent on detailed cost breakdowns.

Simulated interest: 60/100

Vedang Agarwalneutral

What would stop me? Right now, it's the potential cost and whether it integrates smoothly with our existing ComfyUI graphs. We've invested heavily in optimizing our shared machine, and moving to a cloud solution means re-evaluating our entire workflow. The 'pay only for runtime' model sounds interesting, but without specific pricing details on the different GPU tiers – like the H100 or A100 mentioned – it's hard to gauge. We currently run our graphs on-premise, and the upfront hardware cost was significant, but predictable. I'd need to see a clear breakdown of costs before committing to a monthly subscription. Perhaps around $100-$150 per month, depending on the compute resources provided.

Simulated interest: 50/100

Cedric Wickertneutral

The primary concern for me would be the consistency and reliability of the rented GPU instances. Today, we manage our own rented instances, and issues with uptime or performance directly impact our development velocity. While ComfyOnline's promise of 'pay only for runtime' and 'forget about infrastructure' is appealing, I'd need to understand the guarantees around resource availability, especially if we're wrapping stable graphs as HTTP endpoints for demos. The lack of clear pricing tiers beyond 'Get Started For Free' also makes it difficult to estimate costs for our specific, potentially sustained, usage. I would consider paying up to $150 per month, provided it offers reliable performance and scalability.

Simulated interest: 60/100

Primerose Lemaireneutral

My primary concern with trying ComfyOnline this month would be integrating it with our current processes. Today, I meticulously manage all asset versions, approvals, and generation requests through a complex system of spreadsheets and shared drives. While I understand ComfyOnline promises to handle workflows, I need to be assured it can seamlessly integrate or replace these current tracking methods without introducing more manual work or errors. The pricing is also unclear; I see "Pay Only for Runtime," but without specific tiers or examples of usage costs, it's difficult to estimate a monthly budget. If it integrates well and the runtime costs are transparent and manageable, I might consider paying up to $50 per month.

Simulated interest: 40/100

Eduardo Romo Rojasneutral

The main thing stopping me is the manual file transfer I currently do between my local machine and my rented server. If ComfyOnline could seamlessly integrate with my existing cloud storage or offer a better file management solution, I'd consider it. The description mentions "Pay Only for Runtime," which sounds good because I'm frugal and don't want to pay for idle GPUs. However, the pricing structure isn't super clear to me. I'd be willing to pay around $20 per month, maybe up to $30 if it significantly streamlines my workflow and reduces the tedious file copying.

Simulated interest: 60/100

Johanna Veenstraneutral

My current setup works quite well for my needs. I have a robust home desktop for my ComfyUI workflows, and I supplement that with cloud credits when necessary. While the idea of not needing expensive hardware is appealing, I already have it, and the complexity of ComfyUI setup is something I've documented extensively and manage efficiently. My main concern would be the cost-effectiveness for my specific batch variation and upscaling tasks. Without a clear pricing structure for runtime or specific node usage, it's hard to gauge the value proposition compared to my existing cloud credit spend. I'd be willing to pay around $50 per month, but that depends heavily on actual usage and performance.

Simulated interest: 30/100

Alexander Johnsneutral

This looks interesting, but honestly, my current setup is working okay for now. I've gotten pretty good at managing free Colab notebooks and my personal GPU to keep costs down. I'd need to see a clear benefit over what I'm already doing, especially with the asynchronous task queue I just implemented. If I could test it out without needing a credit card upfront, I might give it a whirl. I wouldn't pay more than $20 per month for something like this, and that's if it significantly reduces my current runtime costs and frees up my personal hardware.

Simulated interest: 40/100

Hugh Wolffneutral

My current workflow relies on a robust local setup for ComfyUI, which I manage meticulously. The main hesitation I have with trying ComfyOnline this month is understanding how well it handles custom nodes and specific workflow configurations I've built. We already use a hosted solution for some tasks, but my local fallback is crucial for flexibility and control. I'd need to see clear documentation on custom node support and perhaps a trial period to assess performance before committing. If it proves seamless and as efficient as my current setup, I might consider it, but I'm not ready to pay anything yet. My maximum acceptable price would be $0.

Simulated interest: 30/100

Tanirika Pothuvaalneutral

Honestly, the biggest hurdle for me right now is the ongoing cost. While the idea of not needing expensive hardware is appealing, and the 'pay only for runtime' model sounds great, I'm still very budget-conscious. My current setup, with the recently upgraded graphics card, works well for my needs, and I've optimized my workflow to manage resources efficiently. I would need to see how the runtime costs actually add up for my typical complex projects before committing. For now, I'm trying to stick to free or low-cost environments. If I were to consider it, I might pay around $30 per month, but that depends heavily on the actual usage.

Simulated interest: 40/100

Keep exploring.

Find your next question or see which topics divide AI panels.

Keep exploring

What would you ask next?

Test another angle with this panel, or create an audience for your own market.