Class OpenAIImageEditCapabilityImpl

Description

Provider capability implementation for OpenAIImageEditCapabilityImpl.

Implements

Index

Constructors

Methods

editImage editImageStream

Constructors

constructor

new OpenAIImageEditCapabilityImpl(
provider: BaseProvider,
client: OpenAI,
): OpenAIImageEditCapabilityImpl
Parameters
- provider: BaseProvider
- client: OpenAI
Returns OpenAIImageEditCapabilityImpl
- Defined in src/providers/openai/capabilities/OpenAIImageEditCapabilityImpl.ts:41

Methods

editImage

editImage(
    request: AIRequest<ClientImageEditRequest>,
    executionContext: MultiModalExecutionContext,
    signal?: AbortSignal,
): Promise<AIResponse<NormalizedImage[]>>
Non-streaming image edit.
Parameters
- request: AIRequest<ClientImageEditRequest>
  Client image edit request
- executionContext: MultiModalExecutionContext
  Timeline & context tracking
- Optionalsignal: AbortSignal
  Optional abort signal
Returns Promise<AIResponse<NormalizedImage[]>>
Normalized images and masks
Implementation of ImageEditCapability.editImage
- Defined in src/providers/openai/capabilities/OpenAIImageEditCapabilityImpl.ts:54

editImageStream

editImageStream(
    request: AIRequest<ClientImageEditRequest>,
    executionContext: MultiModalExecutionContext,
    signal?: AbortSignal,
): AsyncGenerator<AIResponseChunk<NormalizedImage[]>>
Streaming image edit.

Images are yielded incrementally as they complete. Masks are yielded ONCE (with the first image chunk) since they represent input context, not generated outputs.
Parameters
- request: AIRequest<ClientImageEditRequest>
- executionContext: MultiModalExecutionContext
- Optionalsignal: AbortSignal
Returns AsyncGenerator<AIResponseChunk<NormalizedImage[]>>
Implementation of ImageEditStreamCapability.editImageStream
- Defined in src/providers/openai/capabilities/OpenAIImageEditCapabilityImpl.ts:123