Create Chat Completion

Create Completion

Text Generation API

Generate images in response to the given request.

Generate SD1.5

Generate SDXL

Generate SD3

Generate ControlNet 1.5

Generate ControlNet SDXL

Generate videos in response to the given request.

Generate SVD

IP Adapter

Media Generation API

Create Tune

Retrieve Tune

Delete Tune

Cancel Tune

List all tunes owned by the current user.

The max number of results to be shown (limit 100).

The base checkpoint ID for the LoRA tune.

List Tunes

Fine Tuning API

List Assets

Create Asset

Delete Asset

Complete Asset Upload

Retrieve asset with the given owner and name or id.

Retrieve Asset

Asset Library API

Asset model. It is used to represent an asset in asset library.

Asset

Asset data. It shows the type of data for an asset.

Asset description. It is used to provide more information.

Asset id. A unique identifier for the asset.

Asset status. It shows the current status of the asset in asset library.

Asset status info. Provides extra information about the asset.

Asset tenant ID. This is unique for each OctoAI user.

Asset type. It is used to distinguish between different supported assets.

AssetType

Base Engine names supported by asset library.

BaseEngine

Base Engine types supported by asset library.

BaseEngineType

CheckpointData

CompleteAssetUploadResponse

Create asset response. It is used to represent the response of create asset API.

CreateAssetResponse

Unused. Present for compatibility purposes.

CreateAssetResponseTransferApi

Data

DataType

DeleteAssetResponse

FileData

FileExtension

FileFormat

FileStructure

HTTPValidationError

LatentData

ListAssetsResponse

LoraData

NoTransferApi

PresignedUrlTransferApi

Get asset URL. Used to download the asset.

RetrieveAssetResponse

RetrieveAssetResponseTransferApi

Status

StsTransferApi

AWS session token. Used for asset transfer.

S3 endpoint URL. Used for asset transfer.

TextualInversionData

TransferApiType

VAEData

ValidationError

ValidationErrorLocItem

Details

ListTunesResponse

True if it has more items than the returned list.

A LoRA tune model that includes tune configurations.

LoraTune

The base checkpoint used for this LoRA Tune.

Resize training images to the "native" dimensions of the respective engine prior to tuning. Images will be proportinally resized to such that the shorter edge will fit the native dimension. For Stable Diffusion 1.5, this is 512 pixels, and for Stable Diffusion XL, this is 1024 pixels. If the shorter edge is shorter than the native dimension, no scaling up will be performed.

The number of steps used for this LoRA Tune.

LoraTuneCheckpoint

The name of the checkpoint. This is cached from asset lake in case the asset is deleted.

An input file to a LoRA tune with its caption.

LoraTuneFile

The caption used when the LoRA tune was requested.

TextToSpeechLatentTune

The Asset Lake asset ids used for this task.

TextTune

The checkpoint name or ID used as base model for tuning.

The number of epochs used for this Text tune.

The learning rate used for this Text tune.

Tune base class. This class is used to store tune information.

Tune

Tune details. This field is used to store a specific tune type configurations.

The output LoRA IDs, if the task was successful.

The details of the status, only used when the associated task failed.

TuneDetails

TuneResult

Number of assets from tune request that are available.

Tune generated assets. This field is filled only when tune was successful.

TuneStatus

TuneType

The ip adapter method for image generation response.

IPAdapterMethod

The ip adapter mode for image generation response.

IPAdapterMode

The image encoding types available for image generation response.

ImageEncoding

ImageGeneration

The generated image or None if it was removed for safety.

True if the image was removed for safety, False otherwise.

Seed used to generate the image. It can be used to recreate the generation.

Generate one or more images based on the given parameters.

ImageGenerationRequest

Floating-point number represeting how closely to adhere to prompt description. Must be a positive number no greater than 50.0.

[Not supported on SD3] Custom checkpoint to be used during image generation.

[Not supported on SD3] Optionally skip later layers of the text encoder. Higher values lead to more abstract interpretations of the prompt.

[Not supported on SD3 or IPAdapter] ControlNet to be used during image generation.

[Not supported on SD3 or IPAdapter] How strong the effect of the controlnet should be.

[Not supported on SD3 or IPAdapter] If provided, indicates fraction of steps at which to stop applying controlnet. This can be used to sometimes generate better outputs.

[Not supported on SD3 or IPAdapter] Controlnet image encoded in b64 string for guiding image generation. Required for controlnet engines.

[Not supported on SD3 or IPAdapter] Whether to apply automatic ControlNet preprocessing.

Integer representing the height of image to generate. None will default to 512 for SD 1.5 and 1024 for SD3, SDXL, and SSD. Supported resolutions (w,h): SD3={(1536, 640), (768, 1344), (832, 1216), (1344, 768), (1152, 896), (640, 1536), (1216, 832), (896, 1152), (1024, 1024)}, SDXL={(1536, 640), (768, 1344), (832, 1216), (1344, 768), (1152, 896), (640, 1536), (1216, 832), (896, 1152), (1024, 1024)}, SD1.5={(768, 576), (1024, 576), (640, 512), (384, 704), (640, 768), (640, 640), (1024, 768), (1536, 1024), (768, 1024), (576, 448), (1024, 1024), (896, 896), (704, 1216), (512, 512), (448, 576), (832, 512), (512, 704), (576, 768), (1216, 704), (512, 768), (512, 832), (1024, 1536), (576, 1024), (704, 384), (768, 512)}, SSD={(1536, 640), (768, 1344), (832, 1216), (1344, 768), (1152, 896), (640, 1536), (1216, 832), (896, 1152), (1024, 1024)}.

[Not supported on SD3] Floating-point number that defines the fraction of steps to perform with the base model. Used only by SD XL. Must be greater than or equal to 0.0 and less than or equal to 1.0.

Define which encoding process should be applied before returning the generated image(s).

Starting point image encoded in b64 string for Image to Image generation mode.

IP Adapter image encoded in b64 string for guiding image generation. Required for ip adapter engines.

b64 encoded mask image for ip adapter. White area should indicate where to paint.

Define which ip adapter method should be applied during the image generation. Supported modes: full, style, and composition

Define which ip adapter mode should be applied during the image generation. Supported modes: base

How strong the effect of the ip adapter should be.

[Not supported on SD3] A dictionary of LoRAs to apply. LoRAs as keys and their weights (float) as values.

[Not supported on SD3] b64 encoded mask image for inpainting. White area should indicate where to paint.

Text describing image traits to avoid during generation.

[Not supported on SD3] Text with a high level description of things to avoid during generation. Used only by SD XL.

Integer representing how many output images to generate with a single prompt/configuration.

[Not supported on SD3] Whether the request requires outpainting or not. If so, special preprocessing is applied for better results.

Text describing the image content to generate.

[Not supported on SD3] Text with a high-level description of the image to generate. Used only by SD XL.

[Not supported on SD3] Sampler name (also known as 'scheduler') to use during image generation.

Integer number or list of integers representing the seeds of random generators. Fixing random seed is useful when attempting to generate a specific image. Must be greater than 0 and less than 2^32.

Integer repreenting how many steps of diffusion to run. Must be greater than 0 and less than or equal to 200.

Floating-point number indicating how much creative the Image to Image generation mode should be. Must be greater than 0 and less than or equal to 1.0.

[Not supported on SD3] Pre-defined styles used to guide the output image towards a particular style. Pre-defined styles are only supported by SDXL.

[Not supported on SD3] A dictionary of textual inversions to be used during image generation. Textual inversions as keys and trigger words as values.

[Not supported on SD3] A dictionary containing a mapping of trigger words to a list of sample images which demonstrate the desired object or style to transfer.

[Not supported on SD3] Whether to enable and apply the SDXL refiner model to the image generation.

[Not Supported on SD3] Custom VAE to be used during image generation.

Integer representing the width of image to generate. None will default to 512 for SD 1.5 and 1024 for SD3, SDXL, and SSD. Supported resolutions (w,h): SD3={(1536, 640), (768, 1344), (832, 1216), (1344, 768), (1152, 896), (640, 1536), (1216, 832), (896, 1152), (1024, 1024)}, SDXL={(1536, 640), (768, 1344), (832, 1216), (1344, 768), (1152, 896), (640, 1536), (1216, 832), (896, 1152), (1024, 1024)}, SD1.5={(768, 576), (1024, 576), (640, 512), (384, 704), (640, 768), (640, 640), (1024, 768), (1536, 1024), (768, 1024), (576, 448), (1024, 1024), (896, 896), (704, 1216), (512, 512), (448, 576), (832, 512), (512, 704), (576, 768), (1216, 704), (512, 768), (512, 832), (1024, 1536), (576, 1024), (704, 384), (768, 512)}, SSD={(1536, 640), (768, 1344), (832, 1216), (1344, 768), (1152, 896), (640, 1536), (1216, 832), (896, 1152), (1024, 1024)}.

ImageGenerationRequestSeed

ImageGenerationResponse

List of ImageGeneration(s) generated by the request.

Total runtime of the image generation(s).

SDXLStyles

The schedulers available for image generation.

Scheduler

VideoGeneration

True if the video/animation was removed for safety, False otherwise.

Safety score of the video/animation generated.

Seed used to generate the video/animation. It can be used to recreate the generation.

The generated video/animation or None if it was removed for safety.

Integer number or list of integers representing the seeds of random generators.Fixing random seed is useful when attempting to generate a specific video/animation (or set of videos/animations). Must be greater than 0 and less than 2^32.

VideoGenerationRequestSeed

VideoGenerationResponse

Total runtime of the video(s) generation(s).

List of VideoGeneration(s) generated by the request.

ChatCompletionChoice

Number of log probabilities to return per output token.

A chat completion message generated by the model.

ChatCompletionChunk

A list of chat completion choices.Can be more than one if n is greater than 1.

The Unix timestamp (in seconds) of when the chat completion was created.

A unique identifier for the entire chat completion request. Each chunk in the stream has the same ID.

An OpenAPI compatible schema for a chat completion chunk choice.

ChatCompletionChunkChoice

An OpenAPI compatible schema for a chat completion choice delta.

ChatCompletionDelta

The tool calls generated by the model, such as function calls.

An OpenAI API compatible schema for a chat completion response object.

ChatCompletionResponse

A unique identifier for the chat completion.

ChatCompletionResponseFormat

An OpenAI API compatible schema for a single ChatMessage.

ChatMessage

Completion choice for completion requests.

CompletionChoice

The generated text in completion response.

Represents a completion response from the API.
Note: both the streamed and non-streamed response objects
share the same shape (unlike the chat endpoint).

CompletionResponse

The Unix timestamp (in seconds) of when the completion was created.

Controls which (if any) tool is called by the model. none means the model will not call any tool and instead generates a message. auto means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools. Specifying a particular tool via {"type": "function", "function": {"name": "my_function"}} forces the model to call that tool. none is the default when no tools are present. auto is the default if tools are present.

List Assets

Query parameters

Response