VertexAIGeminiGenerator

Generate text using Google Gemini models through Vertex AI.

Deprecation Notice

This integration will be deprecated soon. We recommend using GoogleGenAIChatGenerator instead, which provides unified access to both Gemini Developer API and Vertex AI.

Key Features

• Text generation using Google Gemini models through Vertex AI
• Supports multimodal inputs including text and images
• Streaming support for real-time token-by-token responses
• Designed for text generation, not chat (use GoogleGenAIChatGenerator for chat capabilities)
• Authenticates using Google Cloud Application Default Credentials (ADCs)

Configuration

1. Drag the VertexAIGeminiGenerator component onto the canvas from the Component Library.
2. Click on the component to open the configuration panel.
3. On the General tab:
   1. Enter your GCP project ID. Create a secret with the key GCP_PROJECT_ID. For detailed instructions, see Create Secrets.
   2. Optionally, enter the location. If not set, uses us-central1.
   3. Select a model. For available models, see Vertex AI models.
4. Go to the Advanced tab to configure generation settings, safety settings, system instruction, and streaming.

Connections

VertexAIGeminiGenerator accepts multimodal inputs through its parts input — a list of strings, ByteStream objects, or Part objects. It outputs generated text as replies (a list of strings).

Connect PromptBuilder's prompt output to this component's parts input. Connect the replies output to AnswerBuilder.

Source Code

To check this component's source code, open gemini.py in the Haystack Core Integrations repository.

Usage Examples

Basic Configuration

  VertexAIGeminiGenerator:
    type: haystack_integrations.components.generators.google_vertex.gemini.VertexAIGeminiGenerator
    init_parameters:
      model: gemini-2.0-flash

This query pipeline uses VertexAIGeminiGenerator to generate text responses:

components:
  bm25_retriever:
    type: haystack_integrations.components.retrievers.opensearch.bm25_retriever.OpenSearchBM25Retriever
    init_parameters:
      document_store:
        type: haystack_integrations.document_stores.opensearch.document_store.OpenSearchDocumentStore
        init_parameters:
          hosts:
          index: 'default'
          max_chunk_bytes: 104857600
          embedding_dim: 768
          return_embedding: false
          method:
          mappings:
          settings:
          create_index: true
          http_auth:
          use_ssl:
          verify_certs:
          timeout:
      top_k: 10
      fuzziness: 0

  PromptBuilder:
    type: haystack.components.builders.prompt_builder.PromptBuilder
    init_parameters:
      template: |
        Given the following information, answer the question.

        Context:
        {% for document in documents %}
        {{ document.content }}
        {% endfor %}

        Question: {{ query }}
      required_variables:
      variables:

  VertexAIGeminiGenerator:
    type: haystack_integrations.components.generators.google_vertex.gemini.VertexAIGeminiGenerator
    init_parameters:
      project_id:
      model: gemini-2.0-flash
      location:
      generation_config:
      safety_settings:
      system_instruction:
      streaming_callback:

  AnswerBuilder:
    type: haystack.components.builders.answer_builder.AnswerBuilder
    init_parameters:
      pattern:
      reference_pattern:

connections:
  - sender: bm25_retriever.documents
    receiver: PromptBuilder.documents
  - sender: PromptBuilder.prompt
    receiver: VertexAIGeminiGenerator.parts
  - sender: VertexAIGeminiGenerator.replies
    receiver: AnswerBuilder.replies
  - sender: bm25_retriever.documents
    receiver: AnswerBuilder.documents

inputs:
  query:
    - bm25_retriever.query
    - PromptBuilder.query
    - AnswerBuilder.query

outputs:
  answers: AnswerBuilder.answers

max_runs_per_component: 100

metadata: {}

Parameters

Inputs

Parameter	Type	Default	Description
parts	Variadic[Union[str, ByteStream, Part]]		Prompt for the model.
streaming_callback	Optional[Callable[[StreamingChunk], None]]	None	A callback function that is called when a new token is received from the stream.

Outputs

Parameter	Type	Description
replies	List[str]	A list of generated content.

Init Parameters

These are the parameters you can configure in Pipeline Builder:

Parameter	Type	Default	Description
project_id	Optional[str]	None	ID of the GCP project to use. By default, it is set during Google Cloud authentication.
model	str	gemini-2.0-flash	Name of the model to use. For available models, see Vertex AI models.
location	Optional[str]	None	The default location to use when making API calls. If not set, uses us-central-1.
generation_config	Optional[Union[GenerationConfig, Dict[str, Any]]]	None	The generation config to use. Accepted fields: temperature, top_p, top_k, candidate_count, max_output_tokens, stop_sequences.
safety_settings	Optional[Dict[HarmCategory, HarmBlockThreshold]]	None	The safety settings to use.
system_instruction	Optional[Union[str, ByteStream, Part]]	None	Default system instruction to use for generating content.
streaming_callback	Optional[Callable[[StreamingChunk], None]]	None	A callback function that is called when a new token is received from the stream.

Run Method Parameters

These are the parameters you can configure for the component's run() method. This means you can pass these parameters at query time through the API, in Playground, or when running a job. For details, see Modify Pipeline Parameters at Query Time.

Parameter	Type	Default	Description
parts	Variadic[Union[str, ByteStream, Part]]		Prompt for the model.
streaming_callback	Optional[Callable[[StreamingChunk], None]]	None	A callback function that is called when a new token is received from the stream.

Related Information

• Create Secrets
• GoogleGenAIChatGenerator