Embedding Model

The AI model that converts text into vector embeddings for search. Different embedding models have different accuracy, speed, and cost tradeoffs.

Technical definition

Embedding models are neural networks trained on large text corpora to produce dense vector representations. Popular models include OpenAI's text-embedding-3-small (1536 dimensions), Cohere's embed-v3, and open-source alternatives like BGE and E5. The choice of embedding model affects retrieval quality, vector storage size, and inference cost. Newer models generally produce higher quality embeddings at lower dimensions.

How Verabase uses Embedding Model

Verabase uses your LLM provider's embedding model (via BYOK) to generate vector representations of your knowledge base content. By default, OpenAI's text-embedding-3-small is used for its balance of accuracy and cost. When you add or update a source document, Verabase automatically re-embeds the affected chunks. The embedding model choice is configurable in Settings → AI → Embedding Model.

Why it matters for support teams

The embedding model is the foundation of your AI's search quality. A better embedding model means more accurate retrieval, which means better answers. But embedding costs are separate from generation costs — you pay per token embedded. For most teams, the default model (text-embedding-3-small) offers the best balance of quality, speed, and cost.

Related terms

How Verabase Handles Self-Healing Knowledge

The standard industry approach relies on support agents manually flagging stale content and technical writers updating documentation weeks later. Verabase completely upends this model through an autonomous AI platform that identifies knowledge gaps the moment an AI agent encounters a question it cannot answer.

Rather than requiring manual intervention, our visual RAG engine drafts a proposed fix. It analyzes the context of the user's question, reviews your existing knowledge base for conflicts, and creates a clear, structured article or snippet. All you need to do is click 'Approve'. This ensures your AI agents continually get smarter over time without adding to your support team's workload.

Security First Architecture

Many legacy platforms bolt AI onto their existing infrastructure, creating potential data leaks between tenants or exposing sensitive internal documents to end-users. Verabase is built from the ground up with a Bring Your Own Key (BYOK) architecture, enterprise-grade access controls, and strict semantic boundaries. This means your private engineering documents never accidentally leak into customer-facing agent responses.