Configuration¶

EmbedCache is configured through environment variables. You can set these directly or use a .env file.

Configuration File¶

Create a .env file in the same directory where you run EmbedCache:

# Server Configuration
SERVER_HOST=127.0.0.1
SERVER_PORT=8081

# Database Configuration
DB_PATH=cache.db
DB_JOURNAL_MODE=wal

# Embedding Models (comma-separated list)
ENABLED_MODELS=AllMiniLML6V2,BGESmallENV15

# LLM Configuration (optional)
# LLM_PROVIDER=ollama
# LLM_MODEL=llama3
# LLM_BASE_URL=http://localhost:11434
# LLM_API_KEY=
# LLM_TIMEOUT=60

Environment Variables¶

Server Settings¶

Variable	Default	Description
`SERVER_HOST`	`127.0.0.1`	Server bind address
`SERVER_PORT`	`8081`	Server port

Database Settings¶

Variable	Default	Description
`DB_PATH`	`cache.db`	SQLite database file path
`DB_JOURNAL_MODE`	`wal`	SQLite journal mode (`wal`, `truncate`, `persist`)

Model Settings¶

Variable	Default	Description
`ENABLED_MODELS`	`AllMiniLML6V2`	Comma-separated list of enabled embedding models

LLM Settings (Optional)¶

These settings enable LLM-based chunking strategies.

Variable	Default	Description
`LLM_PROVIDER`	(none)	LLM provider: `ollama`, `openai`, `anthropic`
`LLM_MODEL`	`llama3`	Model name to use
`LLM_BASE_URL`	(varies)	API base URL
`LLM_API_KEY`	(none)	API key (required for OpenAI/Anthropic)
`LLM_TIMEOUT`	`60`	Request timeout in seconds

Example Configurations¶

Minimal Configuration¶

# Just use defaults
ENABLED_MODELS=AllMiniLML6V2

Production Configuration¶

SERVER_HOST=0.0.0.0
SERVER_PORT=8080
DB_PATH=/var/lib/embedcache/cache.db
DB_JOURNAL_MODE=wal
ENABLED_MODELS=BGESmallENV15,BGEBaseENV15,MultilingualE5Small

With Ollama LLM¶

SERVER_HOST=127.0.0.1
SERVER_PORT=8081
DB_PATH=cache.db
ENABLED_MODELS=BGESmallENV15
LLM_PROVIDER=ollama
LLM_MODEL=llama3
LLM_BASE_URL=http://localhost:11434

With OpenAI¶

SERVER_HOST=127.0.0.1
SERVER_PORT=8081
DB_PATH=cache.db
ENABLED_MODELS=BGESmallENV15
LLM_PROVIDER=openai
LLM_MODEL=gpt-4o-mini
LLM_API_KEY=sk-your-api-key-here

SQLite Journal Modes¶

Mode	Description	Best For
`wal`	Write-Ahead Logging	High concurrency (recommended)
`truncate`	Truncate journal on commit	Single-process access
`persist`	Don't delete journal	Systems with slow file deletion

Default Request Configuration¶

When API requests don't specify a configuration, these defaults are used:

{
  "chunking_type": "words",
  "chunking_size": 512,
  "embedding_model": "BGESmallENV15"
}