Pricing and top-ups

Echo Token Station is pay as you go: what you use is deducted from the balance you topped up. There is no monthly fee and no setup fee.

How prices are calculated

For each request, the price is calculated with the following formula and deducted from your balance.

Price = input tokens × input unit price + output tokens × output unit price

• A token is the unit in which a model counts text. The text you send (input) and the model's answer (output) are counted separately
• Unit price = the provider's official price × the model's multiplier. The official price is the price each provider publishes in US dollars, excluding tax
• How US dollars are converted to yen: [ ]
• Premium member discount: 12.5% (compared with the Basic member price)
• Cache reads and writes (the lower price when you send the same input repeatedly) are charged in the same categories as each provider's pricing
• Tokens used for reasoning (the model thinking before it answers) are counted as output tokens
• Unit prices are per 1 million tokens and are listed on each model's page in the model list and on the pricing page

Unit prices for each model and a worked example of the price of one request are on the pricing page.

Echo's prices include tax [ ]. Official prices (US dollars) exclude tax.

Video generation prices

Video generation models are charged in the following units. Unit prices are listed on each model's page in the model list.

Because generating a video takes time, the price is settled in two stages: at submission and when generation finishes.

1. At submission: an estimated price, calculated from the requested length and resolution, is deducted from your balance. If your balance or the key's spending limit does not cover the estimate, the submission is refused (HTTP 403)
2. When generation finishes: the price is recalculated from the actual usage the provider reports (the actual number of tokens for Seedance; the actual length, number of reference images and seconds of reference video for MiniMax-H3), and the difference is refunded or additionally deducted. An additional deduction is made even if your balance is not enough
3. When generation fails, or the task has not completed 24 hours after submission: the price deducted at submission is refunded in full

The steps and an example are in Video generation.

Top-ups

• Amount: choose from ¥1,000, ¥3,000, ¥10,000 (tax included)
• Payment method: credit card. Accepted cards: [ ]
• To top up, click 「チャージする」 (Top up) in the console, choose an amount and pay on the Stripe payment page. When the payment is done, you return to the page you started from. The steps are in "Balance and top-ups" in API keys and usage
• After you pay, it can take a few seconds to about a minute before your balance is updated
• Sign-up bonus: [ ]

Expiry and order of use

• The balance with the nearest expiry is used first
• A balance past its expiry date lapses and can no longer be used
• Whether topping up again extends the expiry of your earlier balance: [ ]
• You can see the breakdown of your balance by expiry in 「内訳」 (Breakdown) under 「残高とチャージ」 in the console (shown after the member site Echo AI Hub opens)

When your balance runs out

When your balance is used up, requests are stopped (HTTP 403, Insufficient balance). However, the price of requests can exceed the balance at that time. The excess makes the balance negative and is deducted from your next top-up.

This can happen because the balance is checked against an estimated price before a request is sent, and the price is fixed by the actual token counts after the answer ends. If you send several requests at the same time, each of them passes the check, so the excess can be more than one request. To keep it small, send fewer requests at the same time.

Once a top-up is reflected in your balance, you can use the API again. How to set how much each key can spend is in Limits and rate limits.

Price of failed or interrupted requests

The "upstream API" is the API of the provider that actually runs the model.

Invoices and receipts

• Invoice payment for businesses and qualified invoices (適格請求書) are in preparation. Registration number: [ ]
• Receipts: open them from 「チャージ履歴」 (Top-up history) under 「残高とチャージ」 in the console (shown after the member site Echo AI Hub opens)

Refunds

Refunds: [ ]

Price changes

When a provider lowers its official price, Echo lowers its unit price too. Price increases are announced in the changelog [ ] days in advance.

Sources