Pricing
Priced on usage, never on people.
Six products, bought and priced separately. Every plan states what it includes and what going past it costs, so you can work out the bill before you sign up.
None of the six charges per person.
Every number on this page is a work in progress.
A marketplace has two sides, so it has two price lists: plans if you publish data, prepaid packs if you use it. Every price is the whole price — a request is a request.
Publishing data
Plans are flat rent on capacity. Traffic is metered from the first request, your price is yours, and we take none of it.
| Cove $5/mo |
Most publishers start hereInlet $25/mo |
Bay $125/mo |
Gulf $625/mo |
|
|---|---|---|---|---|
| Parking1 | ||||
| Capacitytables, views, indices | 1 GB | 10 GB | 100 GB | 1 TB |
| Datasets publishedno listing fee | As many as fit | As many as fit | As many as fit | As many as fit |
| Metered2 | ||||
| Traffic served | $1 per 5,000 | $1 per 5,000 | $1 per 5,000 | $1 per 5,000 |
| AI Builder | Your key, free — or AI credits at cost × 1.2 | Your key, free — or AI credits at cost × 1.2 | Your key, free — or AI credits at cost × 1.2 | Your key, free — or AI credits at cost × 1.2 |
| Commission on your price | 0% | 0% | 0% | 0% |
| Service | ||||
| Your price | Anything at or above $0.20 per 1,000 requests, changed with 30 days notice | |||
| Payouts | Monthly, once your balance reaches $10 — or accumulate until you say | |||
| Statement | Earnings − traffic − parking − the rail's transfer fee at cost, itemized and never marked up | |||
| Indian tax | 1% deducted at source (Section 194-O), deposited against your PAN | |||
| Guardrails3 | ||||
| Response size | 512 KB | 512 KB | 512 KB | 512 KB |
| Headroom over your tier | 10% for a week | 10% for a week | 10% for a week | 10% for a week |
| If you stop paying | 30 days read-only grace, with notice to your subscribers — parking is netted from earnings first | |||
| Included4 | ||||
| Zero-code toolchain | Auto API from your schema, Custom API shaped visually, AI Builder from a plain-language sentence | |||
| Hosting | Replicated storage in the region you pick — Germany, Finland or India — plus compute, network, caching, generated docs, billing and payout rails, audit logging, monitoring, uptime | |||
| Reporting | Live traffic and earnings per endpoint | |||
| Cove | Inlet | Bay | Gulf | |
- Parking. one flat rent, one rule: 10× the data, 5× the price — it holds at every step of the ladder. It is not a disk fee; everything under Included is in it. Sustained growth means a tier up, never a surprise charge.
- Metered. the only rows that can cost more than the plan price. Traffic is billed from the first request and netted off your earnings.
- Guardrails. ceilings the platform enforces — never billed
- Included. in every tier, from $5 up
A worked month: park 5 GB (Inlet, $25), price at $1 per 100 requests, serve 50,000 requests. Earnings $500 − meter $10 − parking $25 − 1% TDS $5 − rail fee $1.29 = $458.71 transferred. Our total share is $35 of rent and meter; our share of your price is zero — the marketplace we replaced took 20 to 30%.
Using data
No plan to pick. Start free, then load a prepaid wallet that works across the whole catalogue — so a bill can never exceed what you chose to load.
| Free $0 |
Most people start hereStarter $10 |
Builder $50 |
Pro $100 |
|
|---|---|---|---|---|
| What you get1 | ||||
| Credits in the wallet | — | $10 | $51 | $102 |
| Bonus | — | — | +2% | +2% |
| Requests per minute | 10 | 30 | 60 | 100 |
| Requests per hour | 600 | 1,800 | 3,600 | 6,000 |
| Requests per day | 14,400 | 43,200 | 86,400 | 144,000 |
| Max Requestswallet at $1 per 1,000 requests | — | 10,000 | 51,000 | 102,000 |
| Free Quotaevery day, every plan, on the datasets we host | 100 | 100 | 100 | 100 |
| Valid for | Forever — the free tier never expires | 12 months | 12 months | 12 months |
| Card required | No | — | — | — |
| Rates2 | ||||
| API Bay datasetsour curated catalogue | $1 per 1,000 requests — any endpoint, any query | |||
| Community datasetsfrom independent publishers | The publisher's listed price, shown on the endpoint page and in every response header | |||
| Heavy queries | Same price as a lookup — a request is a request | |||
| Service | ||||
| Wallet | One balance across every dataset on API Bay | |||
| Paying by UPI | ₹ in equals credits out, 1:1, zero fees | |||
| Paying by card | The card network's fee is its own line at checkout, never marked up or hidden | |||
| Guardrails3 | ||||
| Overage bills | None, by construction — spending stops at the wallet balance | |||
| Spending caps | Per dataset, if you want tighter control | |||
| Auto-refill | Off by default; obeys a monthly ceiling you set | |||
| Price increases | A publisher must give 30 days notice before one reaches you | |||
| Response size | 512 KB a response; paginate beyond that | |||
| Start free | Load $10 | Load $50 | Load $100 | |
- What you get. credits, not a subscription — you are charged again only when you load more.
- Rates. $10 = 10,000 calls on any API Bay dataset: sanctions screening, company registries, mandi prices, court judgments, climate history — same rate everywhere.
- Guardrails. ceilings the platform enforces — never billed
Prices in USD, exclusive of applicable taxes (GST invoiced at purchase for Indian customers). Neither side has plan tiers to outgrow: publishing is rent plus a meter, using data is a wallet.
Ten personas answering only from what you give them. Two price lists: parking, for the knowledge we keep ready; answering, a prepaid balance for the requests it serves. Nothing is billed in arrears.
Parking
Your knowledge, kept ready to answer. Paid in advance, monthly or yearly. Each plan holds 10× the knowledge of the one before at 5× the price, so every step up is a 50% discount per gigabyte. Every plan answers 500 requests a month; beyond that, requests come from the balance.
| Free $0 |
Most teams start hereStarter $49/mo |
Growth $245/mo |
Business $1,225/mo |
Dedicated $6,900/mo |
Enterprise By quote |
|
|---|---|---|---|---|---|---|
| Parking1 | ||||||
| Knowledgetext we index from pages, files, pasted text and Minimal tables | 1 MB | 1 GB | 10 GB | 100 GB | 100 GB | More |
| Price per GB | — | $49 | $24.50 | $12.25 | — | — |
| Discount per GBover the plan before | — | — | 50% | 50% | — | — |
| A year, one month free | — | $539 | $2,695 | $13,475 | $75,900 | — |
| Answering | 500 requests a month on every plan, then from the balance below | |||||
| Servers | Shared | Shared | Shared | Shared | One server reserved for you | Several, with standby |
| Guardrails2 | ||||||
| Requests a secondwhole workspace | 1 | 5 | 20 | 50 | 100 | Negotiated |
| IndexingMB of text an hour | 1 | 100 | 500 | 2,000 | 2,000 | Negotiated |
| Knowledge size | The plan's size — the crawler stops there, nothing is deleted | |||||
| Reply time limit | 30 seconds, then the contact form — not charged | |||||
| Requests per visitor a day | 100 | |||||
| Service | ||||||
| Uptime | 99.5% a month, with the cloud provider's credit table: 10% under 99.5%, 30% under 99%, 100% under 95% | Standby server, per order form | ||||
| Support | Community | Priority | Priority | Priority, named contact | Named contact, DPA | |
| Included3 | ||||||
| Team members | 1 | 5 | 25 | 100 | 200 | — |
| Assistants and sites | 1 | 5 | 25 | 50 | 100 | Unlimited |
| Personas | All ten | All ten | All ten | All ten | All ten | All ten |
| Alert channels | 1 | 5 | Unlimited | Unlimited | Unlimited | Unlimited |
| “Powered by” badge | On | Removable | Removable | Removable | Removable | Removable |
| Start free | Starter | Growth | Business | Dedicated | Talk to us | |
- Parking. one flat rent on the size of the text we extract and index; the workspace shows the number as it indexes, and nothing else is counted. Yearly is one month free.
- Guardrails. ceilings the server enforces — never billed. Requests a second are a share of one server; above 50 a second means Dedicated. Indexing runs from one shared queue, paid plans ahead of Free.
- Included. in every plan, from Free up
Answering
A prepaid balance: 500 requests per $1, plus 2% on every load. One visitor message with a reply is one request. When the balance is empty the assistant shows the contact form — no error, no charge.
| Free $0 |
Most people start hereLoad $10 $10 |
Load $50 $50 |
Load $100 $100 |
|
|---|---|---|---|---|
| What you get1 | ||||
| Requests a monthalways, on every plan | 500 | 500 | 500 | 500 |
| Loaded500 per $1 | — | 5,000 | 25,000 | 50,000 |
| Bonus+2% on the load | — | +100 | +500 | +1,000 |
| In the balanceloaded + bonus, spent after the monthly 500 | — | 5,100 | 25,500 | 51,000 |
| A request | — | $0.00196 | $0.00196 | $0.00196 |
| Valid for | Every month | 12 months | 12 months | 12 months |
| Rates2 | ||||
| Larger amounts | In $100 steps, +2% on each | |||
| What counts | One visitor message with a reply. A reply that says the knowledge has no answer and offers the contact form counts. A reply that times out, a reply while the balance is empty, widget loads, the greeting and blocked traffic never count | |||
| AI on your own key | Your provider bills you at its price, for answering and for vectorising your knowledge | |||
| AI credits, no key | Not offered yet — this page will update when it is | |||
| Guardrails3 | ||||
| Overage bills | None, by construction — when the balance is empty the assistant shows the contact form | |||
| Auto top-up | Not offered yet — coming January 2027 | |||
| Alerts, by email | At 20% of the balance and at zero; at 80% of the monthly 500 | |||
| Meter you can check | One row per charged request in the conversation log, with its ID; the balance statement lists requests by day | |||
| Start free | Load $10 | Load $50 | Load $100 | |
- What you get. credits, not a subscription — you are charged again only when you load more. Every plan, Free included, answers 500 requests a month before the balance is touched. A load adds 500 requests per $1 plus 2%, so $10 puts 5,100 in the balance. The 500 return every month; the balance is spent only after them, so a $10 load lasts until 5,100 requests beyond the monthly 500 have been answered.
- Rates. $1 = 500 requests, whichever persona answers and whichever model you bring. The model's own cost is outside: you bring your own key.
- Guardrails. ceilings the platform enforces — never billed
A worked month: a 1,000-page site with 20 documents is about 9.4 MB of text and fits Starter ($49); the first 500 visitor requests are in the plan and the other 14,500 come from the balance at $28.43. $77.43, plus your model's bill on your own key. The same site on DocsBot is $147.
Form-built agents over the database you already run, gated by identity and audited to the call.
| Free $0 |
Most teams start herePro $99/mo |
Scale $299/mo |
Enterprise Talk to us |
|
|---|---|---|---|---|
| Metered1 | ||||
| Database storage | 2 GB | 25 GB | 150 GB | Custom |
| Network in and out | 10 GB | 250 GB | 1500 GB | Custom |
| API requests | 100,000 | 10 million | 50 million | Custom |
| AI tokens | 100,000 | 2 million | 5 million | Custom |
| Agent-hours | 10 hours | 300 hours | 750 hours | Custom |
| Service | ||||
| Audit depth | Full depth | Full depth | Full depth | Full depth |
| Uptime | 95% target | 99.9% target | 99.5% SLA | Negotiated |
| If we miss it | — | No credits — an objective, not an SLA | Service credits, sole remedy | Contractual, per order form |
| Deployment | Hosted | Hosted | Hosted | Hosted or on-premise, air-gapped on-premise |
| Also | Never paused, community support | Access Map | Point-in-time recovery, priority support | The installation holds metadata only, a named contact |
| Guardrails2 | ||||
| Longest a run may take | 5 min | 30 min | 4 hours | Unlimited |
| CPU per agent | 25% | 100% | 150% | Unlimited |
| Memory per agent | 256 MB | 2 GB | 8 GB | Unlimited |
| Agents running at once | 2 | 10 | 50 | Unlimited |
| Included3 | ||||
| Users | 2 | Unlimited | Unlimited | Unlimited |
| Agents | 2 | 25 | 200 | Unlimited |
| Workflows | 1 | 25 | 200 | Unlimited |
| Agent teams | Unlimited | Unlimited | Unlimited | Unlimited |
| MCP tools | 3 | 50 | Unlimited | Unlimited |
| Projects | 1 | 10 | 50 | Unlimited |
| Spaces per projectplus the release workspace | 1 | 5 | 25 | Unlimited |
| Start free | Pro | Scale | Talk to us | |
- Metered. the only rows that can cost more than the plan price. On a paid plan you carry on and pay: $0.80 a GB storage · $0.05 a GB network · $1 a million API requests · $0.05 an agent-hour · $9 a million AI tokens, or free with your own key. On Free we hold no card, so the allowance stops at its limit.
- Guardrails. ceilings the kernel enforces — never billed
- Included. hard limits — you upgrade, you are never billed a surprise
Storage past your allowance costs $0.80 a gigabyte. A plain database charges about $0.125, so this is the one place we are the expensive option. Agent-hours are metered too, at $0.05 an hour, but nobody is charged per person: a five-person team pays what one person pays, and that team pays $250 to $500 a month elsewhere.
Web, mobile and desktop apps over fifteen databases and five file formats, versioned system-wide.
| Free $0 |
Most teams start herePro $59/mo |
Scale $249/mo |
Enterprise Talk to us |
|
|---|---|---|---|---|
| Metered1 | ||||
| Database storage | 1 GB | 25 GB | 150 GB | Custom |
| Network in and out | 5 GB | 250 GB | 1500 GB | Custom |
| API requests | 500,000 | 25 million | 100 million | Custom |
| AI tokens | 50,000 | 1 million | 3 million | Custom |
| Compute-hours | 5 hours | 100 hours | 300 hours | Custom |
| Service | ||||
| Uptime | 95% target | 99.9% target | 99.5% SLA | Negotiated |
| If we miss it | — | No credits — an objective, not an SLA | Service credits, sole remedy | Contractual, per order form |
| Deployment | Hosted | Hosted | Hosted | Hosted or on-premise, air-gapped on-premise |
| Also | Never paused, community support | No compute size to pick, no active-user billing, system-wide versioning and rollback | Point-in-time recovery, read replicas, 30-day backups, priority support | Connect your own databases, SSO and SAML, a DPA |
| Guardrails2 | ||||
| Longest a request or job may take | 10 sec | 1 min | 5 min | Unlimited |
| CPU per instance | 25% | 100% | 150% | Unlimited |
| Memory per instance | 256 MB | 1 GB | 4 GB | Unlimited |
| Executions at once | 5 | 50 | 250 | Unlimited |
| Included3 | ||||
| Users | 2 | Unlimited | Unlimited | Unlimited |
| Appsweb, mobile, desktop or site | 1 | 10 | 100 | Unlimited |
| MCP tools | 3 | 50 | Unlimited | Unlimited |
| Projects | 1 | 10 | 50 | Unlimited |
| Spaces per projectplus the release workspace | 1 | 5 | 25 | Unlimited |
| Start free | Pro | Scale | Talk to us | |
- Metered. the only rows that can cost more than the plan price. On a paid plan you carry on and pay: $0.80 a GB storage · $0.05 a GB network · $1 a million API requests · $0.05 a compute-hour · $9 a million AI tokens, or free with your own key. On Free we hold no card, so the allowance stops at its limit.
- Guardrails. ceilings the kernel enforces — never billed
- Included. hard limits — you upgrade, you are never billed a surprise
Storage past your allowance costs $0.80 a gigabyte. A plain database charges about $0.125, so this is the one place we are the expensive option. The bill follows the meters alone: seats, compute size and monthly active users are not charged. A growing app pays $131 to $645 a month elsewhere.
Flat files into a warehouse we host, questions in plain English, dashboards anyone can open.
| Free $0 |
Most teams start herePro $79/mo |
Scale $199/mo |
Enterprise Talk to us |
|
|---|---|---|---|---|
| Metered1 | ||||
| Database storage | 1 GB | 25 GB | 150 GB | Custom |
| Network in and out | 5 GB | 150 GB | 1000 GB | Custom |
| API requests | 50,000 | 5 million | 25 million | Custom |
| AI tokens | 50,000 | 2 million | 5 million | Custom |
| Dashboard and AI queries | 5,000 | 250,000 | 1 million | Custom |
| Service | ||||
| Uptime | 95% target | 99.9% target | 99.5% SLA | Negotiated |
| If we miss it | — | No credits — an objective, not an SLA | Service credits, sole remedy | Contractual, per order form |
| Deployment | Hosted | Hosted | Hosted | Hosted or on-premise, air-gapped on-premise |
| Also | Bring your own model key, community support | Unlimited datasets, email support | Embedded dashboards, SSO, scheduled reports, priority support | SOC 2, a DPA, a named contact |
| Guardrails2 | ||||
| Longest a query may run | 15 sec | 2 min | 10 min | Unlimited |
| CPU per query | 50% | 200% | 400% | Unlimited |
| Memory per query | 512 MB | 4 GB | 16 GB | Unlimited |
| Queries at once | 2 | 10 | 40 | Unlimited |
| Fastest dashboard refresh | 5 min | 1 min | 15 sec | Unlimited |
| Included3 | ||||
| Usersviewers are never licensed | 2 | Unlimited | Unlimited | Unlimited |
| Dashboards | 1 | Unlimited | Unlimited | Unlimited |
| Saved data personas | 1 | Unlimited | Unlimited | Unlimited |
| Warehousehosted by us, not one you supply | Included | Included | Included | Included |
| MCP tools | 3 | 25 | Unlimited | Unlimited |
| Projects | 1 | 10 | 50 | Unlimited |
| Spaces per projectplus the release workspace | 1 | 5 | 25 | Unlimited |
| Start free | Pro | Scale | Talk to us | |
- Metered. the only rows that can cost more than the plan price. On a paid plan you carry on and pay: $0.80 a GB storage · $0.05 a GB network · $1 a million API requests · $5 per 100,000 queries · $9 a million AI tokens, or free with your own key. On Free we hold no card, so the allowance stops at its limit.
- Guardrails. ceilings the kernel enforces — never billed
- Included. hard limits — you upgrade, you are never billed a surprise
Storage past your allowance costs $0.80 a gigabyte. A plain database charges about $0.125, so this is the one place we are the expensive option. Reading is free: nobody is licensed per viewer, and the warehouse is included rather than something you buy elsewhere and connect. Fifty people reading dashboards costs $370 to $500 a month elsewhere, almost all of it viewer licences.
Software you run on your own hardware, priced entirely on usage: a flat licence, prepaid requests, and support that is free unless you want an engineer on call. Nothing is bundled.
The price list
Four lines, each billed on its own. Every feature is unlocked from the first day, and capacity is unlimited: your hardware is the limit.
| Licence $100/yr |
The only usage meterRequests $1 per 1,000 |
Support $0 |
On-call $150/hr |
|
|---|---|---|---|---|
| What you get1 | ||||
| Covers | Every feature, today and every one we ship later | Calls to Minimal, REST and MCP counted alike | Discord, Slack and email, no time limit | An engineer from Littlebit Minimal on your incident |
| Capacity | Unlimited | Prepaid | Every customer | When you choose it |
| How it is billed2 | ||||
| Billed | Once a year | Prepaid, in packs of $100 and its multiples | Never | Per hour, per engineer |
| Expires | Each year | Never | — | — |
| Response | — | — | Can be slow, unless the Community joins the thread | Immediate |
| Get a licence | ||||
- What you get. the licence unlocks the whole product, so there are no tiers and no feature behind a higher price.
- How it is billed. requests are the only thing that drains. A topup stays valid until you spend it.
Loading requests
Prepaid packs of $100 and its multiples, so a bill can never exceed what you chose to load. Put in a year of usage to see the cost.
Volume discount. If your annual top-ups reach $70,000, talk to us. Discounted request pricing is offered against a minimum annual contract of $70,000.
On-call support
Free support is always there. When you want an answer now, Littlebit Minimal developers join the incident at $150 an hour per engineer. It usually starts with one. A hand-off to another engineer is still one. If an incident needs several engineers at once, we say so and give the reason, and you decide whether to take it.
A worked year
A team serving 10 million requests: licence $100 + 100 packs of $100 (10 million requests, $10,000) + one 2-hour incident with one engineer ($300) = $10,400. Writing the REST API and MCP server yourself takes a team of three senior engineers about nine months, $140,000 (3 × $62,000 × 0.75 year, rounded), and the same team, $186,000 a year (3 × $62,000), to keep it current and add to it. The first year saves $129,600, and each year after saves $175,600.
Terms
By taking a Minimal Core licence you agree not to build products that compete with Littlebit Labs or its associates, unless you hold written exclusive permission to do so.
Try the product first on the hosted free tier at selectstar.wtf.
Every plan on this page is sold on the Terms of Service, and personal data is handled under the Privacy Policy. The Terms define what each quota means, what an overage is billed at, the availability we commit to, and what happens to a project you stop using. The Policy sets out what we hold about you, what stays yours inside a project, and how long each is kept.
We never charge per person.
Users are unlimited on every paid plan, so a team of fifty pays what a team of two pays. Free is capped at two as an evaluation limit. Viewers on a dashboard are never licensed.
Free is never paused.
A free workspace stays up whether it was used yesterday or in March. It answers at once. Go a month untouched and we flag it, changing nothing; at four months we archive it to a portable export; the archive is deleted at sixteen. Six emails across that run, each naming the date of the next step, and the export is one click at every point.
Auth stays with you.
Minimal does not issue your users' credentials and does not connect your third-party services. Both stay where they already are.
The short answers.
What does using data cost?
The first 100 requests a day are free, forever. After that you load a prepaid wallet at $1 per 1,000 requests, and it works across the whole catalogue. Spending stops at what you loaded, so there is never an overage bill.
What does publishing cost?
Flat rent on capacity, from Cove at 1 GB for $5 up to Gulf at 1 TB for $625; ten times the data costs five times the price. Traffic is metered on top at $1 per 5,000 requests, from the first request.
What do I earn from a dataset?
You set the price, any amount from $0.20 per 1,000 requests, and you keep all of it. We take 0% commission.
Is there an enterprise tier, or plans to outgrow?
No. Publishing is rent plus a meter and using data is a wallet, so the pricing is the same for everyone.
How is Chat Studio priced?
In two parts. Parking is the knowledge we keep ready, paid in advance monthly or yearly; each plan holds ten times the knowledge of the one before at five times the price. Answering is a prepaid balance.
What counts as a request?
One visitor message with a reply. A balance buys 500 requests per $1, plus 2% on every load, whichever persona answers and whichever model you bring. Every plan, Free included, answers 500 requests a month before the balance is touched.
What happens when the balance runs out?
Spending stops at what you loaded, so an overage cannot happen. You are charged again only when you load more.
Does the price include the AI model?
No. The model's own cost is outside, on your own key. A 1,000-page site with 20 documents fits Starter at $49, and with 14,500 requests from the balance the month is $77.43 plus your model's bill.
How is Enterprise priced?
By quote, as several servers with standby, since Chat Studio has no database of its own.
What does the licence cover?
$100 a year unlocks every feature, today and every one we ship later. There are no tiers, and capacity is unlimited: your hardware is the limit.
What happens when I run out of requests?
You are throttled, never billed. A topup in packs of $100 buys requests at $1 per 1,000, and a topup stays valid until you spend it.
What does support cost?
Nothing. Discord, Slack and email are free for good. Replies can be slow until the Community joins the thread. When you want an answer now, an engineer from Littlebit Minimal joins your incident at $150 an hour.
When do I talk to you about price?
When your annual topups reach $70,000. Discounted request pricing is offered against a minimum annual contract of $70,000.
What are the limits of Minimal Core?
Your users' credentials and your third-party services stay where they already are, outside Minimal Core. You may build a product that competes with Littlebit Labs or its associates only with written exclusive permission.
What exactly is metered?
Database storage, network, API requests, AI tokens and agent-hours. Everything under Guardrails is a ceiling the platform enforces and is never billed. People, viewers, compute size and monthly active users are not counted.
What happens when I pass what is included?
On a paid plan you keep going at the published rate: $0.80 a gigabyte of database storage, $0.05 a gigabyte of network, $1 per million API requests, $9 per million AI tokens and $0.05 an agent-hour. On Free the allowance stops and you upgrade when you need more, because we never hold a card. Before either happens we email you.
How is Enterprise priced?
Per database connected. It is the plan that can deploy on-premise as well as hosted, and air-gapped on-premise installations hold metadata only, so your data never leaves your network.
Can I move between plans?
Up or down, whenever. Every allowance is one organisation-wide pool, so adding a project draws on what you already have rather than adding a bill.
What exactly is metered?
Database storage, network, API requests, AI tokens and compute-hours. Everything under Guardrails is a ceiling the platform enforces and is never billed. Compute size and active users are not part of the bill.
What happens when I pass what is included?
On a paid plan you keep going at the published rate: $0.80 a gigabyte of database storage, $0.05 a gigabyte of network, $1 per million API requests, $9 per million AI tokens and $0.05 a compute-hour. On Free the allowance stops and you upgrade when you need more, because we never hold a card. Before either happens we email you.
How is Enterprise priced?
Per database connected, hosted or on-premise, and air-gapped on-premise installations hold metadata only, so your data never leaves your network.
Can I move between plans?
Up or down, whenever. Every allowance is one organisation-wide pool, so adding an app draws on what you already have rather than adding a bill.
What exactly is metered?
Database storage, network, API requests, AI tokens and the work Ask Studio does: dashboard queries and AI queries. Everything under Guardrails is a ceiling the platform enforces and is never billed. Viewers on a dashboard are never counted, so anyone can open one.
What happens when I pass what is included?
On a paid plan you keep going at the published rate in the Metered table; storage is $0.80 a gigabyte and network $0.05 a gigabyte. On Free the allowance stops and you upgrade when you need more, because we never hold a card. Before either happens we email you.
How is Enterprise priced?
Per database connected, hosted or on-premise, and air-gapped on-premise installations hold metadata only, so your data never leaves your network.
Can I move between plans?
Up or down, whenever. Every allowance is one organisation-wide pool, so adding a dashboard draws on what you already have rather than adding a bill.
$30M of Minimal, granted every year.
Ninety teams a year get the suite on their own hardware, with $333,333 of credit against the licence and twelve months to spend it. Otherwise, the prices above are the prices.