Google - API - verified
Gemini Embedding embeddings
Gemini Embedding is a legacy embedding model from Google, called by API as gemini-embedding-001. It returns 3,072-dimension vectors by default and accepts up to 2,048 input tokens, for text input. No per-1M-token list price is stated for it. Not listed on the Gemini API pricing page today; Vertex AI lists Gemini Embedding input at $0.00015 per 1,000 count online and $0.00012 batch.
LegacyModel id: gemini-embedding-001
- Default dims
- 3,072
- Max input
- 2,048 tok
- Price / 1M
- Not stated
- Inputs
- Text
Dimensions
Flexible output size- 3,072default
- 1,536
- 768
Default output: 3,072 dimensions.
Listed sizes: 3,072, 1,536 and 768.
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
Not listed on the Gemini API pricing page today; Vertex AI lists Gemini Embedding input at $0.00015 per 1,000 count online and $0.00012 batch.
Usage
Copied verbatim from the page linked below. Placeholders such as project ids and API keys are the source's own.
curl "https://generativelanguage.googleapis.com/v1beta/models/gemini-embedding-001:embedContent" \
-H "Content-Type: application/json" \
-H "x-goog-api-key: $GEMINI_API_KEY" \
-d '{
"taskType": "SEMANTIC_SIMILARITY",
"content": {
"parts": [
{
"text": "What is the meaning of life?"
},
{
"text": "How much wood would a woodchuck chuck?"
},
{
"text": "How does the brain work?"
}
]
}
}'Worth knowing
Text-only model that takes a task_type such as RETRIEVAL_DOCUMENT or RETRIEVAL_QUERY and needs manual normalization below 3072 dimensions; the Gemini API deprecations page lists a May 14, 2028 shutdown with gemini-embedding-2 as the replacement.
Released . Multilingual: yes.
FAQ
How many dimensions does Gemini Embedding output?
Default output: 3,072 dimensions. Listed sizes: 3,072, 1,536 and 768. 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 Gemini Embedding?
2,048 tokens per input, per Google's documentation. Longer text has to be chunked before you embed it.
How much does Gemini Embedding cost?
No per-1M-token list price is stated for it. Not listed on the Gemini API pricing page today; Vertex AI lists Gemini Embedding input at $0.00015 per 1,000 count online and $0.00012 batch. Source: https://cloud.google.com/vertex-ai/generative-ai/pricing.
What can Gemini Embedding embed?
Text input, per the provider. It is documented as multilingual.
Should I still use Gemini Embedding?
Google lists Gemini Embedding as an older model. Existing indexes keep working while it is served, but vectors from different models cannot be mixed, so moving to a newer model means re-embedding your corpus.
Sources
Last verified from the pages below.
- Gemini API embeddings guide - ai.google.dev
- Gemini API model page - ai.google.dev
- Gemini API deprecations - ai.google.dev
- Vertex AI text embeddings API - docs.cloud.google.com
- Vertex AI pricing - cloud.google.com
More Google embedding models
Default dimensions, max input tokens and list price per 1M tokens.
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.