Skip to main content
POST
Generate media with AI

Authorizations

Authorization
string
header
required

API key obtained from your Tella account settings

Body

application/json

Start generating media from a prompt

prompt
string
required

What to generate. Up to 4000 characters for an image, 2000 for a video or a sound effect.

Minimum string length: 1
Example:

"A watercolor illustration of a lighthouse at dawn"

type
enum<string>
required

The kind of media to generate

Available options:
image,
video,
sound-effect
Example:

"image"

durationSeconds
integer

For video only: the clip's length in whole seconds, 4-30 on the 2.5 models and 4-15 on the 2.0 models. Defaults to 5. A longer clip costs proportionally more credits and takes longer to generate.

Required range: 4 <= x <= 30
Example:

10

model
enum<string>

Which model to use. For image: sunburst (GPT Image 2.6, default) is the most capable and follows reference images most precisely; flare (GPT Image 2.6) is faster; seedream-5.5 (Seedream 5.5) renders text most accurately and edits a reference image most precisely; seedream-5.0-pro (Seedream 5.0) is good at text in images and at keeping a reference image's layout; seedream-5.0-flash is its faster version. For video, a Seedance model: seedance-2.5 is the default and has the best quality (1080p); seedance-2.5-4k is the same model in native 4K, about 1.6 times the credits and slower; seedance-2.5-flash is close to 2.5's quality, twice as fast and cheaper (720p); seedance-2.0-fast is cheaper still (720p); seedance-2.0-mini is the cheapest (720p). Not accepted for sound-effect.

Available options:
sunburst,
flare,
seedream-5.5,
seedream-5.0-pro,
seedream-5.0-flash,
seedance-2.5,
seedance-2.5-4k,
seedance-2.5-flash,
seedance-2.0-fast,
seedance-2.0-mini
Example:

"flare"

referenceSourceId
string

For image and video only: the sourceId of an image source to guide the generation — from POST /v1/sources (kind: "image") after uploading the bytes, or an image library item's sourceId.

Example:

"su_abc123def456"

Response

Accepted

The generation that was started, plus the weekly allowance for sound effects

generation
object
required

An AI media generation and, once finished, its library item

limit
integer
required

For sound-effect, the caller's weekly AI generation allowance, which depends on the plan. Images and videos are charged in the workspace's credits instead and have no weekly allowance: they report 1000000

Required range: -9007199254740991 <= x <= 9007199254740991
Example:

20

remaining
integer
required

For sound-effect, generations left in the current week, after this one; 1000000 for images and videos

Required range: -9007199254740991 <= x <= 9007199254740991
Example:

19

resetAt
number
required

For sound-effect, when the weekly allowance resets, as a millisecond epoch; 0 for images and videos

Example:

1760000000000