Pricing and access terms

1. Payment model

Rate Cloud provides access to individual models by subscription for 30 calendar days. The right to use the selected model within the established concurrency is paid for, not a token package. The service has no refillable token balance and no per-request charge in rubles. The exact order amount and currency are shown by the server before payment confirmation.

Each paid subscription provides four concurrent slots for one model. Two active subscriptions for the same model provide eight slots in the account's shared pool; subscriptions for different models create independent pools. A slot is occupied during request processing and released upon completion. Keys of the same account share these pools; additional keys do not add slots.

2. Models and current price

The Rate Cloud pricing lineup includes five models: GLM 5.3, GLM 5.3 Flash, DeepSeek V4.1 Flash, DeepSeek V4 Pro and Qwen 3.8 Max. The listed price applies to one subscription to the selected model for 30 calendar days with four parallel slots. Whether a specific model can be purchased depends on its current status in the server catalog and the result of a technical check. In the base configuration, DeepSeek and Qwen are disabled; their presence on the pricing page alone does not enable a purchase.

The price, order composition and sale status are determined by the current server catalog in the account. If the price on this page and in the order differ, the purchase should be stopped and support contacted. An already paid period is not revalued due to a subsequent catalog change.

3. Period, renewal and additional slots

The period begins after a confirmed payment event and provision of access. Renewal preserves the previously paid term; the end date of each period is shown in the account. An additional subscription adds four slots of the selected model only after payment confirmation. Creating an order does not by itself extend the term or increase rights.

Automatic charges and automatic renewal are not currently offered. Upon expiry of all active periods, access to the model ceases; purchase history and available request history are retained within the established retention periods.

4. Request limits

Rate Cloud does not set a separate monthly quota on the total of input and output tokens for an active subscription. This is not a guarantee of a specific volume of computation, speed, uninterrupted model availability, or unlimited size of an individual request. The technical limits of the selected model and provider, four slots per subscription, and an overall limit of 60 initiated requests per minute per account apply. A fifth concurrent request under a single subscription is not queued and is rejected.

For security, key access rules, its expiry, allowed models and IP/CIDR, as well as service readiness checks are applied. A dependency error may temporarily lead to request rejection. The number of tokens used shown in the account is based on actual usage data received; in the absence of such data it is marked as unmeasured, not zero.

5. Free trial access

After confirming the email and linking Telegram, the account may receive one trial period of three hours with one shared parallel slot for supported models. The trial period begins when both conditions are met and does not increase paid slots. Re-registration or a new key does not create a new trial. The current status and end time are displayed in the account.

6. Checkout and refund

The payment provider and public checkout are not yet connected. The order button records the intent to purchase access but does not accept payment or issue a subscription. Do not transfer funds using details sent in messages or outside the confirmed Rate Cloud payment scenario. Once payments are connected, the published Terms of Service и Refund policy.

To clarify the order status and terms, contact via contacts or in Support Telegram.