Pinecone now supports Google's Gemini Embedding 2 model
According to Pinecone: Pinecone now supports Google's Gemini Embedding 2 model within its Documents API, letting teams run semantic search using plain text without managing API keys. The integration handles embedding calls and billing directly, simplifying development work.
According to Pinecone:
Pinecone announced that its Documents API integrated embedding feature now supports Google's Gemini Embedding 2 model, joining the existing lineup of Pinecone-hosted models. The integration allows development teams to send plain text directly for both document writes and queries, removing the need to build and maintain a separate embedding pipeline or manage Google Cloud accounts and API keys.
Under this setup, Pinecone serves Gemini Embedding 2 via Google Vertex AI, automatically handling error retries and billing token usage on the Pinecone invoice at Google's list price. The system automatically inserts model-specific instructions for documents and queries, and supports customizable dimensions and task types like question answering and classification.
Availability is currently limited to serverless Documents API indexes using the console or Python SDK 10.3.0, with text-only support for now. To encourage adoption, Pinecone is offering double the included monthly tokens (20 million instead of 10 million) for the Builder plan through December 31, 2026.
View selection rationale
Importance 53/100
About this brief
- Content type
- News summary
- Source type
- First-party / official source
- Original source
- Pinecone ↗
- Source published
- Process
- AI-assisted discovery, translation, summarisation, classification and scoring; standards and corrections are maintained by the editorial desk.
- Verification
- The original URL and publication metadata are retained. This brief is not independently fact-checked unless explicitly stated.
- Published by
- Polyemergence Editorial Desk · PolyEmergence Pte Ltd.
- Conflict disclosure
- No financial or commercial conflict declared.