Support

If you need help with integration, billing, or your account, start with the documentation and the FAQ. Contact us through the form if you still need assistance.

Documentation

Guidance for authentication, API keys, requests, and error codes.

Contact

If the documentation and FAQ don’t solve it, send us a request through the form. The appropriate team will get back to you.

FAQ

Answers to common questions about getting started, billing, and your account.

Getting started & Account

With DMM AI Router, you can access multiple leading LLM models through a single API.
You can centralize billing and review and analyze usage. 

By using multiple providers and, in the event of an outage, switching to another model or provider,
you can improve availability while keeping the API unified.

Create a DMM AI Router account and purchase credits from the “Billing” page.
Credits are applied to your LLM usage fees, and the cost of each request is deducted from your balance.

Once you've added credits, issue an API key to start using the API.

The email address cannot be changed because it is a permanent identifier for the account.

If you want to use a different address, create a new account and reconfigure your API keys and settings.

You do not need to delete API keys before closing your account.
However, to ensure that access using API keys is stopped, we recommend revoking the API keys in advance. Revocation usually takes effect almost immediately. 

Once you delete your account, you will no longer be able to view or manage the associated API keys.
Also, deleted accounts cannot be restored.

Models & API Connections

You can check the models supported by DMM AI Router on the “Models” page.

Compare input and output pricing, context length, and supported features to choose the model that best suits your needs.

New models are added on an ongoing basis and are available for use as soon as they are released.

If the model you want to use has not been added, please request it through the contact form.

Response time varies depending on the model used, the content and length of the input and output, and the level of congestion.
Please try the model you plan to use to confirm whether its response speed is suitable for your needs.

Requests are sent to the least expensive provider offering the model you specified. If an error occurs, the request automatically switches to another provider offering the same model. If no alternative is available, an error is returned.

For API calls, send the API key issued on the “API keys” page in the HTTP header
“Authorization: Bearer <API key>”.
You can create and revoke API keys on the “API keys” page.

DMM AI Router supports the OpenAI API-compatible /chat/completions endpoint.
Supported models can be called using the same request and response format as the OpenAI API.

By using separate API keys for each environment, you can manage usage limits and usage history individually.

Response speed may vary depending on the model or provider used.

Model availability and pricing may change.
For the latest information, please check the “Models” page.

We do not provide an API for managing API keys or credits.

Pricing & Billing

DMM AI Router uses a prepaid, pay-as-you-go system in which usage fees are deducted from purchased credits.
Usage fees are calculated based on each model's pricing structure and the amount of processing, such as input and output.

You can check the unit price for each model on the "Models" page.
You can check your usage history and usage fees on the "Activity" page.

The total payment amount and cost breakdown when purchasing credits are displayed on the final purchase confirmation screen.

A 5.5% fee is charged when adding credit.
However, if the amount added is less than $14.55, a minimum fee of $0.80 will be charged.

We use credits denominated in U.S. dollars. You can add credits manually, or
set up “Auto-recharge” to automatically add credits when your balance falls below the amount you set.

Once payment is complete, credit purchases cannot be canceled or refunded for any reason. Please keep this in mind before using the service.

We accept credit cards (Visa/Mastercard/JCB/American Express/Diners Club).

You can change or delete your registered credit card information from the “Billing” page.

You can enable or disable it in the “Auto-recharge” settings on the “Billing” page.

Auto-recharge is automatically disabled, and an email is sent. 
Until you resolve the issue with your payment method or recharge limit and re-enable auto-recharge, 
it will not resume.

Please contact your credit card company.
If necessary, consider paying with a different card.

Credits are valid for 180 days from the date they are added. Unused credits expire after the expiration date and can no longer be used.Adding more credits does not extend the expiration date of existing credits.

Credits with the closest expiration date are used first.

We do not send email notifications before credits expire. 
Please be aware of the credit expiration date when using them.

Usage & Settings

As a general rule, no credits are deducted when a request fails.
You are charged only for requests that are successfully processed.

You can check it in the transaction history on the “Billing” page.

You can download it in CSV format by clicking the “CSV” button on the “Activity” page.

You can set a monthly usage limit for each API key.

Even if you have a credit balance, you may be unable to use the API for the following reasons:

・The API key is invalid
・You have reached the API key usage limit
・The connection settings are incorrect

First, check the key status, amount used, and usage limit on the “API keys” page. 
The API key usage limit cannot be changed after the key is issued. 
To increase the limit, issue a new API key.

Security & Data

Basic request metadata (timestamp, model used, and token count) is recorded.
Prompts and generated results are not recorded, including when an error occurs, because you are opted out by default. 
You can opt in to recording prompts and generated results and change the retention period from the “Settings” page.

However, the retention period varies by model, so please check the “Models” page.

Please check the “Models” page to see whether usage data is used for training by the model provider.

Change your account login password, then revoke any API keys suspected of being leaked.

Also check your usage history, and if you find any suspicious usage, include details such as the date, time, and amount, and
contact us through the contact form.

Troubleshooting & Support

From the contact form on the “Support” page, please provide the issue you are experiencing and the following information, to the extent you know it.
When contacting support, please enter the email address registered with DMM AI Router.

・Date and time of occurrence and time zone 
・Name of the API key used, model name, application, and connection environment 
・Error code
・Error message 
・The steps taken before the issue occurred and any troubleshooting you have already tried 

For inquiries about payments, please also provide the date and time of the transaction and the amount.