Class ChatMistralAI<CallOptions>

Integration with a chat model.

Type Parameters

Hierarchy (view full)

Implements

Constructors

constructor

Properties

apiKey maxTokens model modelName safeMode safePrompt streamUsage streaming temperature topP endpoint? randomSeed? seed?

Methods

bindTools completionWithRetry getLsParams invocationParams withStructuredOutput

Constructors

constructor

Properties

apiKey

apiKey: string

The API key to use.

Default

{process.env.MISTRAL_API_KEY}

maxTokens

maxTokens: number

The maximum number of tokens to generate in the completion. The token count of your prompt plus max_tokens cannot exceed the model's context length.

model

model: string = "mistral-small-latest"

The name of the model to use.

Default

{"mistral-small-latest"}

modelName

modelName: string = "mistral-small-latest"

The name of the model to use. Alias for model

Default

{"mistral-small-latest"}

safeMode

safeMode: boolean = false

Deprecated

use safePrompt instead

safePrompt

safePrompt: boolean = false

Whether to inject a safety prompt before all conversations.

Default

{false}

streamUsage

streamUsage: boolean = true

Whether or not to include token usage in the stream.

Default

{true}

streaming

streaming: boolean = false

Whether or not to stream the response.

Default

{false}

temperature

temperature: number = 0.7

What sampling temperature to use, between 0.0 and 2.0. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.

Default

{0.7}

topP

topP: number = 1

Nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. Should be between 0 and 1.

Default

{1}

Optionalendpoint

endpoint?: string

Override the default endpoint.

OptionalrandomSeed

randomSeed?: number

The seed to use for random sampling. If set, different calls will generate deterministic results. Alias for seed

Optionalseed

seed?: number

The seed to use for random sampling. If set, different calls will generate deterministic results.

Methods

bindTools

completionWithRetry

  • completionWithRetry(input, streaming): Promise<AsyncGenerator<ChatCompletionResponseChunk, any, unknown>>

  • Calls the MistralAI API with retry logic in case of failures.

    Parameters

    • input: ChatRequest

      The input to send to the MistralAI API.

    • streaming: true

    Returns Promise<AsyncGenerator<ChatCompletionResponseChunk, any, unknown>>

    The response from the MistralAI API.

  • completionWithRetry(input, streaming): Promise<ChatCompletionResponse>

  • Parameters

    • input: ChatRequest
    • streaming: false

    Returns Promise<ChatCompletionResponse>

getLsParams

invocationParams

withStructuredOutput

  • withStructuredOutput<RunOutput>(outputSchema, config?): Runnable<BaseLanguageModelInput, RunOutput>

  • Type Parameters

    • RunOutput extends Record<string, any> = Record<string, any>

    Parameters

    • outputSchema: any
    • Optionalconfig: any

    Returns Runnable<BaseLanguageModelInput, RunOutput>

  • withStructuredOutput<RunOutput>(outputSchema, config?): Runnable<BaseLanguageModelInput, {
        parsed: RunOutput;
        raw: BaseMessage;
    }>

  • Type Parameters

    • RunOutput extends Record<string, any> = Record<string, any>

    Parameters

    • outputSchema: any
    • Optionalconfig: any

    Returns Runnable<BaseLanguageModelInput, {
        parsed: RunOutput;
        raw: BaseMessage;
    }>

Settings

Member Visibility

On This Page

Constructors
Properties
Methods