Skip to main content
Watch: I Built an AI App Without Leaving Databricks

Google - API - verified

Multimodal Embeddings (multimodalembedding@001) embeddings

Multimodal Embeddings (multimodalembedding@001) is a current embedding model from Google, called by API as multimodalembedding@001. It returns 1,408-dimension vectors by default and accepts up to 32 input tokens, for text, image and video input. No per-1M-token list price is stated for it. Vertex AI prices text at $0.0002 per 1,000 count, images at $0.0001 each, and video at $0.0005 to $0.002 per count depending on tier (Essential, Standard, Plus).

CurrentModel id: multimodalembedding@001

Default dims
1,408
Max input
32 tok
Price / 1M
Not stated
Inputs
Text, Image, Video

Dimensions

Flexible output size
  • 1,408default
  • 512
  • 256
  • 128

Default output: 1,408 dimensions.

Listed sizes: 1,408, 512, 256 and 128.

Google documents shorter output vectors for this model (Matryoshka-style), so you can store smaller embeddings by asking for fewer dimensions instead of re-embedding with another model.

Pricing

Per 1M input tokens
Not stated
Batch, per 1M
Not stated

Vertex AI prices text at $0.0002 per 1,000 count, images at $0.0001 each, and video at $0.0005 to $0.002 per count depending on tier (Essential, Standard, Plus).

Source: cloud.google.com/vertex-ai/generative-ai/pricing

Usage

Copied verbatim from the page linked below. Placeholders such as project ids and API keys are the source's own.

HTTP request
POST https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT_ID/locations/LOCATION/publishers/google/models/multimodalembedding@001:predict

{
  "instances": [
    {
      "image": {
        "gcsUri": "IMAGE_URI"
      },
      "text": "TEXT"
    }
  ],
  "parameters": {
    "dimension": EMBEDDING_DIMENSION
  }
}

Source: docs.cloud.google.com/gemini-enterprise-agent-platform/models/embeddings/get-multimodal-embeddings

Worth knowing

Vertex AI only model whose image and text vectors share one space for cross-modal retrieval, with English text capped at 32 tokens and video audio ignored.

Release date not stated in the sources. Multilingual: no.

FAQ

How many dimensions does Multimodal Embeddings (multimodalembedding@001) output?

Default output: 1,408 dimensions. Listed sizes: 1,408, 512, 256 and 128. Google documents shorter output vectors for this model (Matryoshka-style), so you can store smaller embeddings by asking for fewer dimensions instead of re-embedding with another model.

What is the max input length of Multimodal Embeddings (multimodalembedding@001)?

32 tokens per input, per Google's documentation. Longer text has to be chunked before you embed it.

How much does Multimodal Embeddings (multimodalembedding@001) cost?

No per-1M-token list price is stated for it. Vertex AI prices text at $0.0002 per 1,000 count, images at $0.0001 each, and video at $0.0005 to $0.002 per count depending on tier (Essential, Standard, Plus). Source: https://cloud.google.com/vertex-ai/generative-ai/pricing.

What can Multimodal Embeddings (multimodalembedding@001) embed?

Text, image and video input, per the provider. It is not documented as multilingual.

Sources

Last verified from the pages below.

More Google embedding models

Default dimensions, max input tokens and list price per 1M tokens.

Google API modelsPrices per 1M tokens, context windows and specs for every Google model in the API registry.

All embedding models

Newsletter

New models, every week

One email per week. New model releases, embedding and retrieval updates and what they mean for what you build.

Keep exploring

Model API pricing and specs

First-party prices per 1M tokens for chat and reasoning models.

Local models

Open-weight LLMs you can download and run, with sizes and licenses.

What is RAG?

Where embeddings fit in a retrieval pipeline, end to end.

Vector databases compared

Where the vectors go once you have picked a model.