SpeqForgePay for what you use.That is the whole pricing page.
No subscription, no tiers, no seats. You get every feature from the first minute, plus free credits each month that run on our own hardware. Our markup is published, every call is itemised, and every charge we make is listed right here. One page, no fine print.
Then $5 of free credits every month after, for as long as you have the account. No card needed to sign up, and free credits run on our own inference hardware. Add funds only if you want the faster hosted models.
Get started freeCommon amounts, or type your own
$10
Enough to see what it does
$25
A working month for most solo users
$50
A small team shipping steadily
$100
Heavy research and app builds
One to one. $25 buys $25 of balance. We take nothing off the top of a top-up.
Included for everyone
- Builds your web, mobile & desktop app
- Feedback analysis & feature specs
- User stories, effort sizing & evidence quotes
- Competitor research & monitoring
- Customer journey mapping
- GTM planning
- Risk register & DFMEA
- Opportunity analysis
- Digital team, mail & workflows
- Public feedback portals
- Product board & cycles
- Architecture & BDD generation
- Whiteboards & documents
- PDF, CSV, TXT & URL ingestion
- Email support
Phone numbers for your digital team are $1 per month each, billed separately. Nothing else recurs.
What that actually buys you
A balance is hard to picture, so here is real work, priced at what it has cost in practice at standard speed.
Find out what your users are actually asking for
Raw feedback in, a ranked and evidence-backed list of what to build next out.
- 20 batches of feedback analysed into specs, each quoting the users it came from
- 3 competitors researched and written up: pricing, positioning, where they are weak
- 10 market and opportunity research runs on the gaps that turns up
$2.71
Covered by your free credits, with change left over.
Get a working app in front of someone
Your work board becomes a real web, mobile, or desktop app you can hand to a user.
- 150 build turns, each one round of the AI writing your app, testing it, and rebuilding
- 10 batches of the feedback that comes back, analysed into the next round of work
$2.70
Covered by your free credits, with change left over.
Run a full quarter
Everything above, at the pace of a team actually shipping, for three months.
- 120 batches of feedback analysed
- 20 competitor deep-dives
- 60 market and opportunity research runs
- 400 build turns across your apps
$21.40
One $25 top-up covers the quarter.
The same numbers, run by run
Every figure above comes out of this. Real work mixes these, so treat each column as “if you spent it all on this”.
| Amount | Feedback batches analysedone pile of feedback turned into a ranked spec | Competitor deep-divesone competitor researched and written up | Market & opportunity research runsone question researched across the web and your own data | App build turnsone round of the AI editing your app and rebuilding it |
|---|---|---|---|---|
| $5Free every month after | 50 | 25 | 120 | 400 |
| $10Free when you sign up | 100 | 55 | 250 | 850 |
| $25A typical top-up | 270 | 140 | 600 | 2,000 |
| $100Heavy research and building | 1,100 | 580 | 2,400 | 8,000 |
Counts come from the median billed amount per completed run on our own hardware and are rounded down. Your own runs will vary with how much you feed them, and the Faster speed uses premium models and costs more per run.
How your balance is spent
Most usage-based products will not tell you their markup. Ours is 5%, and here is the whole calculation, written out so you can check it against the provider’s own price list rather than take our word for it.
Provider cost, plus 5%
You pay the published API price of the model that ran, plus five percent.
Cache savings go to you
When a provider serves part of a request from its own cache at a discount, the discount reaches your balance instead of being pocketed.
Our hardware costs you less
Work served by our own inference hardware is billed at roughly half the cheapest hosted model, because it costs us less to run.
Every call is itemised in your account settings: what ran, what it cost, and which feature spent it.
Everything you can be charged for
Two things. This list is complete.
AI usage
The published API price of the model that ran, plus 5%. Drawn from your free credits first, then your paid balance.
Phone numbers
$1 per number per month, for as long as you keep it. Billed to your card, not your balance. Only if you provision one.
What never appears on your bill
The charges other tools add, that this one does not.
- No subscription, no plan tier, no minimum spend
- No per-seat fee. Invite your whole team
- No feature locked behind an upgrade
- No charge to sign up, and no card asked for
- No expiry on money you have added, and no charge in a month you do not use it
- No fee on a top-up. $25 buys $25 of balance
Run it on your own cloud
A self-hosted deployment puts SpeqForge inside your own AWS, GCP, or Azure environment, with model usage billed against your provider accounts rather than a balance here. Dedicated support and onboarding included.
Talk to usCommon questions
How much does it cost to run?
It depends entirely on what you run. Pasting feedback in and reading the analysis costs a fraction of a cent. Building a full app or running deep research costs more. Your balance and a full per-feature breakdown of where it went are in your account settings, updated after every call.
What is the Faster speed setting?
Account settings has a Speed picker with two options. Standard runs your work on our own hardware and spends your free credits.
Faster routes it to premium models: quicker, stronger on the hardest jobs, and real money per call. So Faster spends your paid balance and never your free credits.
Leave it on Standard and your free credits go first, with paid balance covering anything past them.
Is there a monthly fee?
No. There is no subscription, no plan tier, and no monthly minimum. You get free credits every month, $10 when you sign up and $5 after, and you add your own funds only if you want the Faster speed or run past them. If you do not use it for three months, you are charged nothing for those three months.
How do the free monthly credits work?
Signing up comes with $10, and every month after that with $5. That goes further than it sounds: $5 covers about 50 feedback batches or 25 competitor deep-dives at standard speed.
They arrive whether or not you have ever paid us anything, and they run on our own inference hardware, which is what makes them free to give you and why they cover every feature at standard speed.
Each month resets to that amount rather than accumulating, so an unused balance does not carry over. Anything you top up on your own card is separate, never expires, and is what pays for the Faster speed.
Are any features locked behind a higher tier?
No. Every feature is available to every account from the first minute, including app builds, the digital team, competitor research, and whiteboards. The only thing that limits you is your balance.
How do I know the 5% is really 5%?
Every call is itemised in your account settings with the model that ran, its token counts, and what you were charged. Those are the model provider’s own published per-token prices, so you can check any line against the provider's pricing page and the arithmetic will come out.
What is the markup on AI usage?
Five percent over the model provider’s published API price, the same rate whichever model runs. That margin covers retries and the calls that fail after we have already been billed for the prompt.
When a provider serves part of a request from its own cache at a discount, the discount is passed through to you rather than kept. Work served by our own inference hardware is billed lower again.
Is anything else billed separately?
Phone numbers. Each number you provision for a digital team member is $1 per month for as long as you keep it, billed by card rather than drawn from your balance. Nothing else recurs.
What payment methods do you accept?
All major credit and debit cards via Stripe. Your payment details are never stored on our servers.
Is my data kept private?
Your feedback and research data is used only to generate your output. It is never shared, sold, or used to train models.
Can I run SpeqForge on my own cloud infrastructure?
Yes. A self-hosted deployment puts SpeqForge into your own AWS, GCP, or Azure environment instead of our shared infrastructure, with usage billed against your own provider accounts. Reach out and we will scope it for your team.
$10 of free credits to start
Then $5 free every month. No card, no subscription, every feature unlocked from day one.
Get started free