Models
Vectora works with pluggable AI models for embeddings, reranking, and LLMs. This guide covers supported models and how to customize them.
Vectora’s Standard Stack
| Component | Model | Provider | Docs |
|---|---|---|---|
| LLM | Gemini 3 Flash | Google AI | → Gemini |
| Embeddings | Voyage 4 | Voyage AI | → Voyage |
| Reranker | Voyage Rerank 2.5 | Voyage AI | → Voyage |
Operation Modes
| Feature | BYOK Mode (Free) | Managed Mode (Plus) |
|---|---|---|
| Standard Models | Gemini + Voyage | Gemini + Voyage |
| Key Management | You provide (BYOK) | Managed by Vectora |
| AI Cost | Paid to provider (or Free tier) | Included in Vectora plan |
| Setup | Manual (vectora config) | Automatic (Zero Config) |
| Privacy | Local keys | Managed keys |
Next Readings
- Gemini Configuration — Setup Google AI
- Voyage Configuration — Setup Voyage AI
- Concepts — How embeddings work
Want to use another model? Open an Issue
External Linking
| Concept | Resource | Link |
|---|---|---|
| Voyage Embeddings | Voyage Embeddings Documentation | docs.voyageai.com/docs/embeddings |
| Voyage Reranker | Voyage Reranker API | docs.voyageai.com/docs/reranker |
| Gemini API | Google AI Studio Documentation | ai.google.dev/docs |
Part of the Vectora ecosystem · Open Source (MIT) · Contributors