Commands Overview

LLM Client provides a comprehensive set of commands organized by functionality. Most commands have short aliases for faster usage.

Command Structure

lc [COMMAND] [SUBCOMMAND] [OPTIONS] [ARGS]

Command Categories

Core Commands

Command	Alias	Description
lc "prompt"	-	Send a direct prompt using defaults
lc chat	lc c	Start interactive chat session
lc providers	lc p	Manage LLM providers
lc models	lc m	List and filter available models
lc keys	lc k	Manage API keys
lc config	lc co	Configure defaults
lc logs	lc l	View and manage chat history
lc completions	-	Generate shell completion scripts

Audio Commands

Command	Alias	Description
lc transcribe	lc tr	Convert audio to text
lc tts	-	Convert text to speech

Advanced Commands

Command	Alias	Description
lc embed	lc e	Generate and store embeddings
lc vectors	lc v	Manage vector databases
lc similar	lc s	Search for similar content
lc search	lc se	Web search integration
lc sync	lc sy	Sync configuration to cloud
lc mcp	-	Manage MCP servers
lc alias	lc a	Manage model aliases
lc templates	lc t	Manage templates
lc proxy	lc pr	Run proxy server
lc web-chat-proxy	lc w	Web chat proxy

Direct Prompts

The simplest way to use lc:

# Using defaults
lc "What is the capital of France?"

# Specify model
lc -m openai:gpt-4 "Write a Python function"

# Specify both
lc --provider openrouter -m "claude-3.5-sonnet" "Explain quantum computing"

# With vector database context (RAG)
lc -v knowledge "What do you know about machine learning?"

# With MCP tools
lc -t fetch "What's the latest news about AI?"

# With web search
lc --use-search brave "What are the latest AI developments?"

# With audio attachments
lc "What is being discussed?" --audio meeting.mp3

# Transcribe audio
lc transcribe recording.wav

# Text to speech
lc tts "Hello world" --output greeting.mp3

Global Options

These options work with most commands:

• -p, --provider <PROVIDER> - Specify provider
• -m, --model <MODEL> - Specify model
• -s, --system <SYSTEM_PROMPT> - Set system prompt
• --max-tokens <MAX_TOKENS> - Maximum number of tokens
• --temperature <TEMPERATURE> - Adjust response randomness
• -a, --attach <ATTACHMENTS> - Attach files
• -u, --audio <AUDIO_FILES> - Attach audio files for transcription
• -t, --tools <TOOLS> - Include MCP tools (comma-separated)
• -v, --vectordb <VECTORDB> - Use vector database for context
• -d, --debug - Enable debug mode
• -c, --continue - Continue previous session
• --cid <CHAT_ID> - Specify chat ID
• --use-search <SEARCH> - Use search results as context
• -h, --help - Show help information
• -V, --version - Show version

Command Aliases

LLM Client uses intuitive aliases to speed up your workflow:

Single Letter Aliases

• c → chat
• p → providers
• m → models
• k → keys
• l → logs
• e → embed
• v → vectors
• s → similar

Two Letter Aliases

• co → config
• sy → sync
• se → search
• pr → proxy

Subcommand Aliases

• a → add
• r → remove or recent or refresh
• l → list
• s → show or setup or stats
• u → update
• d → delete or dump
• i → info

Examples

Quick Provider Setup

# Long form
lc providers add openai https://api.openai.com/v1
lc keys add openai
lc providers models openai

# Short form (same result)
lc p a openai https://api.openai.com/v1
lc k a openai
lc p m openai

Chat Workflow

# Start chat
lc c -m gpt-4

# View recent chats
lc l r

# Get last answer
lc l r a

# Extract code from last answer
lc l r a c

Vector Database Workflow

# Create embeddings
lc e -m text-embedding-3-small -v docs "Important information"

# Search similar content
lc s -v docs "related query"

# Use in chat
lc c -v docs -m gpt-4

MCP Tools Workflow

# Add MCP server
lc mcp add fetch "uvx mcp-server-fetch" --type stdio

# List functions
lc mcp functions fetch

# Use in prompt
lc -t fetch "Get current weather in Tokyo"

# Use in chat
lc c -m gpt-4 -t fetch

Search Integration Workflow

# Add search providers (auto-detected from URL)
lc search provider add brave https://api.search.brave.com/res/v1/web/search
lc search provider add ddg https://api.duckduckgo.com/  # Free option!
lc search provider add jina https://s.jina.ai/

# Set API keys (DuckDuckGo doesn't need one)
lc search provider set brave X-Subscription-Token YOUR_API_KEY
lc search provider set jina Authorization YOUR_API_KEY

# Direct search
lc search query brave "latest AI news" -f json
lc search query ddg "free search query" -f json

# Advanced Jina features
lc search provider set jina X-Engine direct  # Enable full content reading
lc search query jina "comprehensive topic" -f json

# Use in prompts
lc --use-search brave "What's happening in AI today?"
lc --use-search ddg "Free search integration test"
lc --use-search jina "Research with full content"

Audio Workflow

# Transcribe audio files
lc transcribe interview.mp3 --format json
lc tr podcast.wav --language en

# Text to speech
lc tts "Welcome message" --voice nova
lc tts --file script.txt --output narration.mp3

# Use audio in chat
lc "Summarize this meeting" --audio meeting_recording.mp3
lc c -m gpt-4 --audio interview.wav

Shell Completions Setup

# Setup completions for your shell
lc completions zsh >> ~/.zshrc      # Zsh
lc completions bash >> ~/.bashrc    # Bash  
lc completions fish > ~/.config/fish/completions/lc.fish  # Fish

# Or use eval for dynamic completions (recommended)
echo 'eval "$(lc completions zsh)"' >> ~/.zshrc
source ~/.zshrc

# Now enjoy tab completion!
lc -p <TAB>           # Shows your configured providers
lc -m <TAB>           # Shows available models
lc providers <TAB>    # Shows: add, remove, list, models

Getting Help

Every command has built-in help:

# General help
lc --help

# Command help
lc providers --help
lc p --help

# Subcommand help
lc providers add --help
lc p a --help

Next Steps

Explore specific command documentation:

• Providers Commands
• Models Commands
• Chat Commands
• Shell Completions
• MCP Commands
• Audio Commands
• Embedding Commands
• Vector Commands
• Search Commands