Embed Command

Generate embeddings for text using various embedding models. The embed command converts text into high-dimensional vectors that can be stored in vector databases for semantic search and RAG applications.

Overview

Text embeddings are essential for building semantic search systems, recommendation engines, and RAG (Retrieval-Augmented Generation) workflows. The embed command supports multiple embedding models and can process both direct text input and files.

Usage

# Generate embeddings for text
lc embed "Your text here"

# Specify model and provider
lc embed -m text-embedding-3-small --provider openai "Important information"

# Store in vector database
lc embed -v knowledge-base "Document content"

# Process files
lc embed -f document.txt,data.pdf

# Using aliases
lc e "Sample text"

Subcommands

The embed command is a standalone command without subcommands. All functionality is controlled through options and arguments.

Options

Short	Long	Description	Default
-m	--model	Embedding model to use	None
-p	--provider	Provider for the embedding model	None
-v	--vectordb	Vector database to store embeddings	None
-f	--files	Files to process (comma-separated)	None
-d	--debug	Enable debug output	False
-h	--help	Print help	False

Examples

Basic Text Embedding

# Simple text embedding
lc embed "Machine learning is transforming software development"

# With specific model
lc embed -m text-embedding-3-large "Complex technical documentation"

Store in Vector Database

# Create embeddings and store in database
lc embed -v docs "Important company policy information"

# Add multiple pieces of content
lc embed -v knowledge "User manual section 1"
lc embed -v knowledge "User manual section 2"
lc embed -v knowledge "FAQ answers"

Process Files

# Single file
lc embed -f documentation.md -v docs

# Multiple files
lc embed -f "manual.pdf,guide.txt,readme.md" -v knowledge

# With specific model
lc embed -m text-embedding-ada-002 -f data.txt -v research

RAG Workflow

# Step 1: Create embeddings from documents
lc embed -f "docs/*.md" --vectordb project-docs

# Step 2: Use in chat with vector context
lc -v project-docs "How do I deploy the application?"

# Step 3: Find similar content
lc similar -v project-docs "deployment process"

See Also

• Vectors Command
• Similar Command

Troubleshooting

Common Issues

"Embedding model not found"

• Error: Specified embedding model doesn't exist
• Solution: Use lc models embed to list available embedding models
• Solution: Check provider supports the model

"File not found"

• Error: Specified file doesn't exist
• Solution: Check file paths and permissions
• Solution: Use absolute paths for clarity

"Vector database error"

• Error: Cannot connect to or create vector database
• Solution: Ensure database name is valid
• Solution: Check disk space and permissions

"Rate limiting"

• Error: Too many embedding requests
• Solution: Add delays between requests
• Solution: Use batch processing for large files

Best Practices

1. Choose appropriate models: Larger models provide better quality but cost more
2. Chunk large texts: Break long documents into smaller sections
3. Consistent models: Use the same embedding model for search and storage
4. Batch processing: Process multiple texts together for efficiency

Embedding Models

# List available embedding models
lc models embed

# Common embedding models:
# - text-embedding-3-small (OpenAI)
# - text-embedding-3-large (OpenAI)
# - text-embedding-ada-002 (OpenAI)
# - embed-english-v3.0 (Cohere)