Skip to main content

Integration FAQ

Demo​

Signature​

API call demo​

Application Issues​

  • Interface endpoint and appId, what are they?

    After creating the application, you can find the application information below.

  • Where can I find the API documentation and instructions?

    Refer to the left sidebar of the Tutorial under the sections: Basic Job, Workflow Job, Resource Management, and Get Model. It provides an introduction to API demos and links.

  • Explanation of Parameters:

    • The image generation parameters include several stages such as INPUT_INITIALIZE, DIFFUSION, etc. Each stage has its necessary parameters, which can be referred to in the corresponding demo.

    • The image upload interface is divided into two steps. In the first step, you obtain the resourceId along with the required address information for the upload image request: putUrl and headers. In the second step, you use this information to initiate the upload request. Upon successful request, you can use the resourceId for image generation or subsequent operations. (Currently, there is no interface available to retrieve uploaded images using resourceId.)

    • The requestId can be generated using the following commonly used method: generating an MD5 hash of the current timestamp and then converting it to hexadecimal format. Here's a Python code example:

      print(hashlib.md5(str(int(time.time())).encode()).hexdigest())

    • Rate Limiting for the interface is currently set to 5 QPS. It can be adjusted according to specific usage requirements in the future.

Computation Issues​

  • What is the price of each computational credit?

    0.003 USD

  • Does image generation failure consume computational credits?

    Nope

  • What is the computational credit estimation API?

    The API is as follows, Input image generation parameters to obtain computational power estimation.

    Get Job Credits

    If using a template, you can use the following interface:

    Workflow Template Check

Payment Issues​

  • Where is the top up entrance, and how can I top up?

    On the Application Management page, click on Info, where you can see the Top Up button as shown in the figure below. Clicking on it allows you to make payments to the corresponding account. After making the payment, click on the Contact button in the bottom right corner to contact customer service for computational credit top up.

  • How to issue an invoice?

    You can contact us to issue an invoice.

Availability​

  • Will there be a situation where image generation is slow during a certain period of time?

    All image generation tasks will enter the image generation queue, and there may be longer waiting times during certain peak periods. Users can avoid these time slots.

  • Sometimes there may be a high failure rate?

    Image generation failures are typically due to inappropriate parameter settings. Users can check and modify certain parameters. For example, setting the Upscaler too high may lead to out of memory and result in image generation failure.

  • Is it possible to exclusively use GPU resources without queuing with others?

    Currently, it is not possible.

Model Issues​

  • Can you provide an interface for listing models?

    Currently not available.

  • How can I obtain models?

    See Get Model Info

  • Can I use my own models for image generation? How do I upload them?

    You can upload models through the Tensor Model Page.

  • Has the previous model been taken down?

    Models that have been taken down cannot be guaranteed for use. For details, please contact us.

Other Issues​

  • Have the generated images been reviewed?

    Nope

  • For API calls, are there any membership restrictions from Tensor?

    TAMS computational credit is calculated separately and is not related to Tensor.

  • Can computational credit from Tensor be used on TAMS?

    They are independent of each other and cannot be used interchangeably.