Class LlmGatewayService

The LLM Gateway service: an OpenAI-compatible chat completions API plus AssemblyAI's Speech Understanding endpoints.

Hierarchy

BaseService
- LlmGatewayService

Index

Constructors

Methods

chatCompletions fetch fetchJson fetchResponse listModels understanding validateUnderstanding

Constructors

constructor

new LlmGatewayService(params): LlmGatewayService
Create a new LLM Gateway service.
Parameters
- params: BaseServiceParams
  The parameters to use for the service.
Returns LlmGatewayService
Overrides BaseService.constructor
- Defined in src/services/llm-gateway/service.ts:21

Methods

chatCompletions

chatCompletions(request): Promise<ChatCompletionResponse>
Create a chat completion.
Parameters
- request: ChatCompletionRequest
  The chat completion request. stream: true is not supported by this client yet.
Returns Promise<ChatCompletionResponse>
Throws
LlmGatewayError when the request fails.
- Defined in src/services/llm-gateway/service.ts:31

`Protected`fetch

fetch(input, init?): Promise<Response>
Parameters
- input: string
- Optionalinit: RequestInit
Returns Promise<Response>
Inherited from BaseService.fetch
- Defined in src/services/base.ts:55

`Protected`fetchJson

fetchJson<T>(input, init?): Promise<T>
Type Parameters
- T
Parameters
- input: string
- Optionalinit: RequestInit
Returns Promise<T>
Inherited from BaseService.fetchJson
- Defined in src/services/base.ts:79

`Protected`fetchResponse

fetchResponse(input, init?): Promise<Response>
Parameters
- input: string
- Optionalinit: RequestInit
Returns Promise<Response>
Inherited from BaseService.fetchResponse
- Defined in src/services/base.ts:22

listModels

listModels(): Promise<ListModelsResponse>
List the models available on the LLM Gateway.

Returns Promise<ListModelsResponse>
Throws
LlmGatewayError when the request fails.
- Defined in src/services/llm-gateway/service.ts:51

understanding

understanding(request): Promise<LlmGatewaySpeechUnderstandingResponse>
Run a Speech Understanding request.
Parameters
- request: LlmGatewaySpeechUnderstandingRequest
Returns Promise<LlmGatewaySpeechUnderstandingResponse>
Throws
LlmGatewayError when the request fails.
- Defined in src/services/llm-gateway/service.ts:63

validateUnderstanding

validateUnderstanding(request): Promise<void>
Validate a Speech Understanding request without running it.
Parameters
- request: LlmGatewaySpeechUnderstandingRequest
Returns Promise<void>
Throws
LlmGatewayError when the request is invalid.
- Defined in src/services/llm-gateway/service.ts:78