Elastic Jina Embeddings v3 Now Available in Gemini Enterprise Agent Platform Model Garden

With Jina Embeddings v3 deployed directly inside their Google Cloud projects and Virtual Private Clouds (VPCs), enterprises can run embedding models within their own infrastructure, avoiding external API calls and maintaining control over sensitive data.

This deployment model supports:

• Total Data Sovereignty: Sensitive data doesn’t leave the security perimeter, making it ideal for regulated industries like the public sector.
• Predictable Pricing: Users pay a flat hourly rate for compute rather than per-token metering, rewarding scale and high-volume traffic.
• Native Integration: Seamlessly connects with existing Google Cloud services, including BigQuery, Cloud Storage, IAM for security, and unified billing.

“We’re bringing Jina models into Model Garden to make high-performance retrieval easier to deploy inside enterprise environments,” said Ken Exner, chief product officer at Elastic. “Organizations are no longer forced to choose between performance and control. They can now run production-ready embeddings directly within their Google Cloud VPC, with full ownership of their data and predictable costs.”

Jina Embeddings v3 supports a range of multilingual semantic search and retrieval use cases and is widely used in production across the Elasticsearch ecosystem. The model is well-suited for applications such as semantic search and RAG pipelines.

The launch of Jina v3 in Model Garden is part of Elastic’s broader rollout to provide a full, end-to-end retrieval stack, with additional embedding and reranking models planned.

For more information, read the blog: Jina embeddings v3 now available on Gemini Enterprise Agent Platform Model Garden

Availability