Class FireworksLLM

Fireworks.ai implementation of the BaseLLM class Fireworks.ai provides an OpenAI-compatible API, so we use the OpenAI SDK with a custom baseURL

Hierarchy (View Summary)

BaseLLM
- FireworksLLM

Constructors

constructor

new FireworksLLM(apiKey: string, baseURL?: string): FireworksLLM
Parameters
- apiKey: string
- OptionalbaseURL: string
Returns FireworksLLM
Overrides BaseLLM.constructor
- Defined in Providers/Fireworks/src/models/fireworks.ts:20

Properties

`Protected`_additionalSettings

_additionalSettings: Record<string, any>

Protected property to store additional provider-specific settings

`Protected`thinkingStreamState

thinkingStreamState: ThinkingStreamState

State tracking for streaming thinking extraction Providers should initialize this if they support thinking models

Accessors

AdditionalSettings

get AdditionalSettings(): Record<string, any>
Get the current additional settings

Returns Record<string, any>
Inherited from BaseLLM.AdditionalSettings
- Defined in Core/dist/generic/baseLLM.d.ts:41

`Protected`apiKey

get apiKey(): string
Only sub-classes can access the API key

Returns string
Inherited from BaseLLM.apiKey
- Defined in Core/dist/generic/baseModel.d.ts:166

Client

get Client(): OpenAI
Read-only getter method to get the client instance

Returns OpenAI
- Defined in Providers/Fireworks/src/models/fireworks.ts:33

SupportsPrefill

get SupportsPrefill(): boolean
Whether this LLM provider supports assistant prefill (pre-seeding the start of the model's response). Providers that support prefill should override this to return true. This is used as a code-level default when database metadata (AIModelType/AIModel/AIModelVendor.SupportsPrefill) is null. Database values of true/false override this getter.

Returns boolean
Inherited from BaseLLM.SupportsPrefill
- Defined in Core/dist/generic/baseLLM.d.ts:88

SupportsStreaming

get SupportsStreaming(): boolean
Fireworks.ai supports streaming

Returns boolean
Overrides BaseLLM.SupportsStreaming
- Defined in Providers/Fireworks/src/models/fireworks.ts:40

Methods

`Protected`addThinkingToMessage

addThinkingToMessage(
message: ChatCompletionMessage,
thinkingContent?: string,
): ChatCompletionMessage
Add thinking content to a chat completion message
Parameters
- message: ChatCompletionMessage
- OptionalthinkingContent: string
Returns ChatCompletionMessage
Inherited from BaseLLM.addThinkingToMessage
- Defined in Core/dist/generic/baseLLM.d.ts:177

ChatCompletion

ChatCompletion(params: ChatParams): Promise<ChatResult>
Process a chat completion request. If streaming is enabled and supported, this will route to the streaming implementation.
Parameters
- params: ChatParams
Returns Promise<ChatResult>
Inherited from BaseLLM.ChatCompletion
- Defined in Core/dist/generic/baseLLM.d.ts:59

ChatCompletions

ChatCompletions(
paramsArray: ChatParams[],
callbacks?: ParallelChatCompletionsCallbacks,
): Promise<ChatResult[]>
Process multiple chat completion requests in parallel. This is useful for:
- Generating multiple variations with different parameters (temperature, etc.)
- Getting multiple responses to compare or select from
- Improving reliability by sending the same request multiple times
Parameters
- paramsArray: ChatParams[]
  Array of chat completion parameter objects
- Optionalcallbacks: ParallelChatCompletionsCallbacks
  Optional callbacks for progress and individual completions
Returns Promise<ChatResult[]>
Promise resolving to an array of ChatResults in the same order as the input params
Inherited from BaseLLM.ChatCompletions
- Defined in Core/dist/generic/baseLLM.d.ts:70

ClassifyText

ClassifyText(params: ClassifyParams): Promise<ClassifyResult>
Parameters
- params: ClassifyParams
Returns Promise<ClassifyResult>
Overrides BaseLLM.ClassifyText
- Defined in Providers/Fireworks/src/models/fireworks.ts:400

ClearAdditionalSettings

ClearAdditionalSettings(): void
Clear all additional settings This is useful for resetting the state of the provider or when switching between different configurations.

Returns void
Inherited from BaseLLM.ClearAdditionalSettings
- Defined in Core/dist/generic/baseLLM.d.ts:54

ConvertMJToOpenAIChatMessages

ConvertMJToOpenAIChatMessages(
messages: ChatMessage[],
): ChatCompletionMessageParam[]
Parameters
- messages: ChatMessage[]
Returns ChatCompletionMessageParam[]
- Defined in Providers/Fireworks/src/models/fireworks.ts:404

ConvertMJToOpenAIRole

ConvertMJToOpenAIRole(role: string): "system" | "user" | "assistant"
Utility method to map a MemberJunction role to OpenAI role
Parameters
- role: string
Returns "system" | "user" | "assistant"
- Defined in Providers/Fireworks/src/models/fireworks.ts:468

`Protected`createStreamingRequest

createStreamingRequest(
params: ChatParams,
): Promise<AsyncIterable<ChatCompletionChunk, any, any>>
Create a streaming request for Fireworks.ai
Parameters
- params: ChatParams
Returns Promise<AsyncIterable<ChatCompletionChunk, any, any>>
Overrides BaseLLM.createStreamingRequest
- Defined in Providers/Fireworks/src/models/fireworks.ts:256

`Protected`extractThinkingFromContent

extractThinkingFromContent(
content: string,
): { content: string; thinking?: string }
Extract thinking content from non-streaming content This method handles case-insensitive extraction of thinking blocks
Parameters
- content: string
Returns { content: string; thinking?: string }
Inherited from BaseLLM.extractThinkingFromContent
- Defined in Core/dist/generic/baseLLM.d.ts:157

`Protected`finalizeStreamingResponse

finalizeStreamingResponse(
    accumulatedContent: string,
    lastChunk: ChatCompletionChunk,
    usage: {
        completionTokens: number;
        promptTokens: number;
        totalTokens: number;
    },
): ChatResult
Create the final response from streaming results for Fireworks.ai
Parameters
- accumulatedContent: string
- lastChunk: ChatCompletionChunk
- usage: { completionTokens: number; promptTokens: number; totalTokens: number }
Returns ChatResult
Overrides BaseLLM.finalizeStreamingResponse
- Defined in Providers/Fireworks/src/models/fireworks.ts:341

GetFileCapabilities

GetFileCapabilities(): FileCapabilities
Returns the native file input capabilities of this LLM driver, or null if the driver does not support file attachments. Subclasses that accept files (PDFs, images, etc.) should override this method.

Returns FileCapabilities
Inherited from BaseLLM.GetFileCapabilities
- Defined in Core/dist/generic/baseLLM.d.ts:94

`Protected`getThinkingTagFormat

getThinkingTagFormat(): { close: string; open: string }
Get the thinking tag format for this provider Providers can override this to customize the thinking tag format

Returns { close: string; open: string }
Inherited from BaseLLM.getThinkingTagFormat
- Defined in Core/dist/generic/baseLLM.d.ts:149

`Protected`handleStreamingChatCompletion

handleStreamingChatCompletion(params: ChatParams): Promise<ChatResult>
Template method for handling streaming chat completion This implements the common pattern across providers while delegating provider-specific logic to abstract methods.
Parameters
- params: ChatParams
Returns Promise<ChatResult>
Inherited from BaseLLM.handleStreamingChatCompletion
- Defined in Core/dist/generic/baseLLM.d.ts:100

`Protected`initializeThinkingStreamState

initializeThinkingStreamState(): void
Initialize thinking stream state for streaming extraction

Returns void
Inherited from BaseLLM.initializeThinkingStreamState
- Defined in Core/dist/generic/baseLLM.d.ts:164

`Protected`nonStreamingChatCompletion

nonStreamingChatCompletion(params: ChatParams): Promise<ChatResult>
Implementation of non-streaming chat completion for Fireworks.ai
Parameters
- params: ChatParams
Returns Promise<ChatResult>
Overrides BaseLLM.nonStreamingChatCompletion
- Defined in Providers/Fireworks/src/models/fireworks.ts:133

`Protected`processStreamChunkWithThinking

processStreamChunkWithThinking(rawContent: string): string
Process streaming chunk with thinking extraction This method handles case-insensitive extraction across chunk boundaries
Parameters
- rawContent: string
Returns string
Inherited from BaseLLM.processStreamChunkWithThinking
- Defined in Core/dist/generic/baseLLM.d.ts:173

`Protected`processStreamingChunk

processStreamingChunk(
    chunk: ChatCompletionChunk,
): {
    content: string;
    finishReason?: string;
    usage?: {
        completionTokens: number;
        promptTokens: number;
        totalTokens: number;
    };
}
Process a streaming chunk from Fireworks.ai
Parameters
- chunk: ChatCompletionChunk
Returns {
    content: string;
    finishReason?: string;
    usage?: {
        completionTokens: number;
        promptTokens: number;
        totalTokens: number;
    };
}
Overrides BaseLLM.processStreamingChunk
- Defined in Providers/Fireworks/src/models/fireworks.ts:315

`Protected`resetStreamingState

resetStreamingState(): void
Clear per-request streaming state. Invoked by the base class at the start and end of every streaming request.

Returns void
Overrides BaseLLM.resetStreamingState
- Defined in Providers/Fireworks/src/models/fireworks.ts:125

`Protected`resetThinkingStreamState

resetThinkingStreamState(): void
Reset thinking stream state

Returns void
Inherited from BaseLLM.resetThinkingStreamState
- Defined in Core/dist/generic/baseLLM.d.ts:168

SetAdditionalSettings

SetAdditionalSettings(settings: Record<string, any>): void
Set additional provider-specific settings Subclasses should override this method to validate required settings
Parameters
- settings: Record<string, any>
  Provider-specific settings
Returns void
Inherited from BaseLLM.SetAdditionalSettings
- Defined in Core/dist/generic/baseLLM.d.ts:48

SummarizeText

SummarizeText(params: SummarizeParams): Promise<SummarizeResult>
Parameters
- params: SummarizeParams
Returns Promise<SummarizeResult>
Overrides BaseLLM.SummarizeText
- Defined in Providers/Fireworks/src/models/fireworks.ts:382

`Protected`supportsThinkingModels

supportsThinkingModels(): boolean
Check if the provider supports thinking models Providers should override this to return true if they support thinking extraction

Returns boolean
Inherited from BaseLLM.supportsThinkingModels
- Defined in Core/dist/generic/baseLLM.d.ts:144

Class FireworksLLM

Hierarchy (View Summary)

Index

Constructors

Properties

Accessors

Methods

Constructors

constructor

Parameters

Returns FireworksLLM

Properties

Protected_additionalSettings

ProtectedthinkingStreamState

Accessors

AdditionalSettings

Returns Record<string, any>

ProtectedapiKey

Returns string

Client

Returns OpenAI

SupportsPrefill

Returns boolean

SupportsStreaming

Returns boolean

Methods

ProtectedaddThinkingToMessage

Parameters

Returns ChatCompletionMessage

ChatCompletion

Parameters

Returns Promise<ChatResult>

ChatCompletions

Parameters

Returns Promise<ChatResult[]>

ClassifyText

Parameters

Returns Promise<ClassifyResult>

ClearAdditionalSettings

Returns void

ConvertMJToOpenAIChatMessages

Parameters

Returns ChatCompletionMessageParam[]

ConvertMJToOpenAIRole

Parameters

Returns "system" | "user" | "assistant"

ProtectedcreateStreamingRequest

Parameters

Returns Promise<AsyncIterable<ChatCompletionChunk, any, any>>

ProtectedextractThinkingFromContent

Parameters

Returns { content: string; thinking?: string }

ProtectedfinalizeStreamingResponse

Parameters

Returns ChatResult

GetFileCapabilities

Returns FileCapabilities

ProtectedgetThinkingTagFormat

Returns { close: string; open: string }

ProtectedhandleStreamingChatCompletion

Parameters

Returns Promise<ChatResult>

ProtectedinitializeThinkingStreamState

Returns void

ProtectednonStreamingChatCompletion

Parameters

Returns Promise<ChatResult>

ProtectedprocessStreamChunkWithThinking

Parameters

Returns string

ProtectedprocessStreamingChunk

Parameters

Returns { content: string; finishReason?: string; usage?: { completionTokens: number; promptTokens: number; totalTokens: number; };}

ProtectedresetStreamingState

Returns void

ProtectedresetThinkingStreamState

Returns void

SetAdditionalSettings

Parameters

Returns void

`Protected`_additionalSettings

`Protected`thinkingStreamState

`Protected`apiKey

`Protected`addThinkingToMessage

`Protected`createStreamingRequest

`Protected`extractThinkingFromContent

`Protected`finalizeStreamingResponse

`Protected`getThinkingTagFormat

`Protected`handleStreamingChatCompletion

`Protected`initializeThinkingStreamState

`Protected`nonStreamingChatCompletion

`Protected`processStreamChunkWithThinking

`Protected`processStreamingChunk

Returns {
content: string;
finishReason?: string;
usage?: {
completionTokens: number;
promptTokens: number;
totalTokens: number;
};
}

`Protected`resetStreamingState

`Protected`resetThinkingStreamState

`Protected`supportsThinkingModels