Class ChatModelIntegrationTests<CallOptions, OutputMessageType, ConstructorArgs>Abstract

constructor

• new ChatModelIntegrationTests<CallOptions, OutputMessageType, ConstructorArgs>(fields): ChatModelIntegrationTests<CallOptions, OutputMessageType, ConstructorArgs>

• Type Parameters

  • CallOptions extends BaseChatModelCallOptions = BaseChatModelCallOptions
  • OutputMessageType extends BaseMessageChunk = BaseMessageChunk
  • ConstructorArgs extends RecordStringAny = RecordStringAny

  Parameters

  • fields: ChatModelIntegrationTestsFields<CallOptions, OutputMessageType, ConstructorArgs>

  Returns ChatModelIntegrationTests<CallOptions, OutputMessageType, ConstructorArgs>

functionId

invokeResponseType

runTests

• runTests(): Promise<boolean>

• Run all unit tests for the chat model. Each test is wrapped in a try/catch block to prevent the entire test suite from failing. If a test fails, the error is logged to the console, and the test suite continues.

  Returns Promise<boolean>

testBatch

• testBatch(callOptions?): Promise<void>

• Parameters

  • OptionalcallOptions: any

  Returns Promise<void>

testBindToolsWithOpenAIFormattedTools

• testBindToolsWithOpenAIFormattedTools(): Promise<void>

• Returns Promise<void>

testBindToolsWithRunnableToolLike

• testBindToolsWithRunnableToolLike(): Promise<void>

• Returns Promise<void>

testCacheComplexMessageTypes

• testCacheComplexMessageTypes(): Promise<void>

• Returns Promise<void>

testConversation

• testConversation(callOptions?): Promise<void>

• Parameters

  • OptionalcallOptions: any

  Returns Promise<void>

testInvoke

• testInvoke(callOptions?): Promise<void>

• Parameters

  • OptionalcallOptions: any

  Returns Promise<void>

testInvokeMoreComplexTools

• testInvokeMoreComplexTools(): Promise<void>

• Tests a more complex tool schema than standard tool tests. This schema contains the Zod field: z.record(z.unknown()) which represents an object with unknown/any fields. Some APIs (e.g Google) do not accept JSON schemas where the object fields are unknown.

  Returns Promise<void>

testModelCanUseToolUseAIMessage

• testModelCanUseToolUseAIMessage(): Promise<void>

• This test verifies models can invoke a tool, and use the AIMessage with the tool call in a followup request. This is useful when building agents, or other pipelines that invoke tools.

  Returns Promise<void>

testModelCanUseToolUseAIMessageWithStreaming

• testModelCanUseToolUseAIMessageWithStreaming(): Promise<void>

• Same as the above test, but streaming both model invocations.

  Returns Promise<void>

testStream

• testStream(callOptions?): Promise<void>

• Parameters

  • OptionalcallOptions: any

  Returns Promise<void>

testStreamTokensWithToolCalls

• testStreamTokensWithToolCalls(): Promise<void>

• Returns Promise<void>

testStructuredFewShotExamples

• testStructuredFewShotExamples(callOptions?): Promise<void>

• Test that model can process few-shot examples with tool calls.

  Parameters

  • OptionalcallOptions: any

  Returns Promise<void>

testToolMessageHistoriesListContent

• testToolMessageHistoriesListContent(callOptions?): Promise<void>

• Test that message histories are compatible with list tool contents (e.g. Anthropic).

  Parameters

  • OptionalcallOptions: any

  Returns Promise<void>

testToolMessageHistoriesStringContent

• testToolMessageHistoriesStringContent(callOptions?): Promise<void>

• Test that message histories are compatible with string tool contents (e.g. OpenAI).

  Parameters

  • OptionalcallOptions: any

  Returns Promise<void>

testUsageMetadata

• testUsageMetadata(callOptions?): Promise<void>

• Parameters

  • OptionalcallOptions: any

  Returns Promise<void>

testUsageMetadataStreaming

• testUsageMetadataStreaming(callOptions?): Promise<void>

• Parameters

  • OptionalcallOptions: any

  Returns Promise<void>

testWithStructuredOutput

• testWithStructuredOutput(): Promise<void>

• Returns Promise<void>

testWithStructuredOutputIncludeRaw

• testWithStructuredOutputIncludeRaw(): Promise<void>

• Returns Promise<void>