Skip to content

Plan the path from request to result.

Understand what you would send, how you would follow a video Job and where the result fits into your app.

Integration preview. API access and payments are not open. Examples cannot be run, and details may change.

Text generation

The planned text workflow starts with a prompt and returns generated text. We plan to include usage information with the response.

  1. Prepare the text or prompt you want to send.
  2. Submit a request using the planned API format.
  3. Read the generated response and usage information.
View the illustrative example
PREVIEW ONLY β€” not executable
REQUEST_TEXT(
  endpoint = "https://example.invalid/text",
  credential = "NOT_A_REAL_KEY",
  input = "DEMO_CONTENT_ONLY"
)
READ_RESPONSE_AND_USAGE()

Video generation

We plan to handle video generation as an asynchronous Job. You would check its status and retrieve the result after successful completion. A Job can also fail.

  1. Submit a description to create a video Job.
  2. Check whether the Job is waiting, running or has finished.
  3. Retrieve a result after successful completion and handle failed Jobs.
View the illustrative example
PREVIEW ONLY β€” not executable
CREATE_VIDEO_JOB("DEMO_CONTENT_ONLY")
OBSERVE_JOB_STATUS("JOB_ID_PLACEHOLDER")
IF_COMPLETED: READ_RESULT_WHEN_AVAILABLE()

Credits and model tokens

Credits will measure platform usage. Model tokens measure text sent to or returned by a model; they are a different unit. Pricing and Credit purchases are not available yet.

Talk through your integration

Tell us how text or video generation would fit into your app and what you need to understand before integrating.

Contact t-bridge