Model Settings

HolderChat user guide

Documentation

Summary

The Model tab controls which AI models answer chats, database type, optional BYOK OpenAI key, retrieval limits, and entity-aware RAG routing.

Prerequisites

  • Edit chatbot → Model tab

Settings reference

SettingDefaultRange / options
LLM ModelPlatform defaultList from server
Data Base TypeStandard DatabaseStandard \Enhanced Database
Simple LLM ModelSame familyUsed when simple mode on
Use simple model firstOffCheckbox
Use your own OpenAI API keyOffCheckbox
OpenAI API KeyEmptysk-... when BYOK on
Vector Search Limit103–20
Message History Limit202–200
Entity-aware RAG routingVariesOn = hybrid entity + semantic search

Data Base Type

TypeDescription
Standard DatabaseVector search (default; used by WordPress Connector)
Enhanced DatabaseLightRAG-enhanced knowledge graph path

Changing type may require rebuild and longer processing.

Simple Model Mode

When Use simple model first is enabled:

  1. Lightweight model tries to answer first (lower cost)
  2. Falls back to advanced LLM Model if needed

When disabled, only the advanced LLM Model is used.

API Key Mode (BYOK)

When Use your own OpenAI API key is enabled:

  • Your key is used for this bot’s OpenAI calls
  • You are billed by OpenAI directly
  • Saved key is retained if you temporarily disable BYOK

When disabled, the platform default key is used (included in beta).

Vector Search Limit

Number of knowledge chunks retrieved per user question (3–20). Higher = more context but slower and costlier.

Message History Limit

How many recent user/assistant turns the AI sees (2–200). Full history is still stored in Conversation History.

Entity-aware RAG routing

When enabled, the bot detects entities in questions and uses hybrid search (semantic + entity chunks). When disabled, legacy semantic-only search is used.

What triggers a rebuild

ActionRebuild needed?
Change LLM / simple model settingsNo
Change database typeOften yes — wait for processing
Change vector/history limitsNo
Toggle entity-aware RAGMay require rebuild for index

Limitations

  • Enhanced database not available on all plans (beta: check status messages)
  • Invalid OpenAI key causes chat errors immediately
  • Model list depends on server configuration

Troubleshooting

Symptom: “Invalid API key” errors

Cause: BYOK key wrong or expired.

Fix: Re-enter key; disable BYOK to use platform key.

Symptom: Slow responses

Cause: High vector limit or enhanced DB.

Fix: Lower Vector Search Limit; enable simple model mode.

Related topics