Hoveydai

Choose your plan. The price of every model, before you spend.

Ways to start

Start free, pay for what you use.

Plan details could not be loaded. Sign in to check current plans; example allowances are not a purchase offer.

Free

FreeWith a daily request cap
  • The free models in the list
  • Chat in English
  • No top-up needed
  • Check current limits in your dashboard
Start free

Pay as you go

Table priceInput and output of each model, per 1 million tokens
  • Shared usage credit
  • A separate API key per app
  • Usage and cost per key
  • Access subject to model and plan limits
Create an API key

Paid plans

Check the dashboardCurrent plan details are unavailable here
  • Review included usage credit
  • Check model access and request limits
  • See available payment methods
View billing
Where your message goesIt also reaches the provider of the model you chose.
Deleting a chatDelete archives a chat; it does not erase it.
The numbersEvery price on this page is a sample and is replaced by the live list.
Usage credit

Every request is one line in the ledger.

Your plan pays for chat and models in the app; your wallet pays for the API and API keys. They are separate. These estimates use the catalog token prices; actual usage depends on the request.

  1. Plan and walletChat is covered by your plan; the API is paid from your wallet. They are separate, and each only pays for its own use.
  2. Model priceEach model has two prices: per 1 million tokens you send (input) and per 1 million tokens the model writes (output).
  3. DeductionRequest cost = (input tokens × input price + output tokens × output price) / 1,000,000. It comes off the balance and is recorded against that key.
One request, worked outSample
Modelgpt-5.5
Cost of one request
PartTokensPrice per 1 millionCost
Input8001.250.001
Output600100.006
This request$0.007
Credit and limitsWhen your wallet cannot cover an API request, it is rejected with HTTP 402; in chat, a message is refused once your plan allowance is used up. Refund and credit-expiry policies remain subject to publication.
Cost calculator

One model, one usage pattern, one number.

Enter the size of your own requests to see the cost per request and per month. The chart shows the same usage on every model.

5 to 1,000100

Assumes 30 days a month. Tokens and prices are samples; the API shows the real token count of each request in its response and the usage report.

Monthly cost with this model Sample
21USD
Per request
0.007 USD
Per day
0.7 USD
Per 1,000 requests
7 USD
The same usage on every model, cheapest first (USD per month)
qwen3-coderFree
mistral-small-3.2Free
deepseek-v3.21.428
gpt-5-mini4.2
gemini-3-flash5.22
deepseek-r15.28
claude-haiku-4.511.4
qwen3-max13.68
mistral-large-315.6
gpt-5.521
gemini-3-pro26.4
claude-sonnet-5.534.2
Usage budget preview

Pick an amount, see what it buys.

Use this illustrative USD budget to estimate request volume. Online wallet top-up is currently unavailable. Paid plan checkout, when configured, is available in the billing dashboard.

Example usage budget (USD)
Example starting credit Sample3.5 USD
Top-up10 USD
Example total budget13.5 USD
This buys about 1,428 replies from gpt-5.5 (sample: 800 in, 600 out)
View billing options

Budget preview only. This does not start a payment or add credit.

Receipts

Keep track of payments and usage.

Review payment status and references in your dashboard. The rows below are illustrative, not your payment history. Usage reports are available separately by API key.

Top-up historySample
Sample receipts
DateReceipt no.DescriptionAmount (USD)StatusActions
Sep 29, 2026R-000412Wallet top-up2PaidExample
Sep 17, 2026R-000377Wallet top-up0.5PaidExample
Sep 10, 2026R-000351Wallet top-up1FailedNo receipt issued
Sep 10, 2026R-000352Wallet top-up1PaidExample
Sep 1, 2026R-000298Wallet top-up0.2PaidExample
Top-up receiptFor each successful payment: date, amount, reference number.
Usage reportCost by key and by model, so you know where the money went.
Official invoiceTax and official invoices are not decided yet. Until then this page promises nothing.
Billing questions

What people ask before they top up.

Which currency is used?

The model usage catalog and its estimates use US dollars (USD). English paid-plan checkout uses the published USD plan price when a supported provider is configured. Confirm the amount and currency before paying.

What is a token, and why two prices?

A token is a unit of counted text. What you send (input) and what the model writes (output) are counted and priced separately. The token count of each request appears in the API response.

Does credit expire?

Not decided yet. Expiry and refunds are set before launch and announced on this page.

What happens when the balance runs out?

If your wallet cannot cover the request, the API returns HTTP 402. Charge your wallet in the dashboard. Chat limits are separate and come from your plan.

Is the cost of each key shown separately?

Yes. Usage and cost are shown per key, but there is one API wallet and every key draws from it.

Does a free model use my credit?

No. Models labelled Free cost nothing but have a daily request cap.

Why is every number labelled Sample?

The live list is not available. The numbers show the shape of the page and are replaced by the live list.

Create your first key.

One key. The leading models. Every price in the open.