elith
(v0.1.0)
// configuration [ 0.03 ]

Configuration

Configure Elith for your workflow.

Configuration

Configure Elith through ~/.elith/config.toml and environment variables.

// Configuration File

Elith uses TOML format for configuration. The file is created automatically when you run elith init.

Location: ~/.elith/config.toml

Example Configuration

[default]
model = "claude"

[claude]
api_key = "sk-ant-api03-..."
model = "claude-sonnet-4-20250514"

[lmstudio]
base_url = "http://localhost:1234/v1"
model = "local-model"

[openrouter]
api_key = "sk-or-v1-..."
model = "openai/gpt-4o"

// Interactive Setup

The easiest way to configure Elith:

bashcopy
elith init

This launches an interactive wizard with arrow key navigation:

// flow
Elith Configuration Wizard
Let's set up your AI providers
Available providers:
→ Claude (Anthropic)
Cloud-based, most capable
LM Studio (Local)
Run models locally, privacy-focused
OpenRouter (Multiple models)
Access to multiple providers
Use ↑/↓ arrow keys to navigate, Enter to select

// Environment Variables

Set these in .env in your project root, or export them in your shell:

# Provider API Keys
ANTHROPIC_API_KEY=sk-ant-api03-...
OPENAI_API_KEY=sk-...
GOOGLE_API_KEY=AIza...

# LM Studio Configuration (optional)
LMSTUDIO_BASE_URL=http://localhost:1234/v1

Note: Elith reads .env files automatically via python-dotenv.

// Provider-Specific Configuration

Claude (Anthropic)

[default]
model = "claude"

[claude]
api_key = "sk-ant-api03-..."
model = "claude-sonnet-4-20250514"

Available Models:

  • claude-sonnet-4-20250514 (recommended)
  • claude-3-5-sonnet-20241022
  • claude-3-opus-20240229

Get API Key: https://console.anthropic.com/

LM Studio (Local)

[default]
model = "lmstudio"

[lmstudio]
base_url = "http://localhost:1234/v1"
model = "local-model"

Setup:

  1. Download LM Studio: https://lmstudio.ai/
  2. Load a model (e.g., CodeLlama, Mistral, Llama)
  3. Start the server (default port: 1234)
  4. Configure Elith to use lmstudio

No API key required — runs entirely locally.

OpenRouter

[default]
model = "openrouter"

[openrouter]
api_key = "sk-or-v1-..."
model = "openai/gpt-4o"

Available Models:

  • openai/gpt-4o
  • openai/gpt-3.5-turbo
  • anthropic/claude-3.5-sonnet
  • meta-llama/llama-3.1-8b-instruct
  • google/gemini-pro-1.5

Get API Key: https://openrouter.ai/keys

// Switching Providers

To switch providers, run elith init again or manually edit ~/.elith/config.toml:

[default]
model = "lmstudio"  # Change this to switch providers

Then restart the backend:

bashcopy
elith service restart

// Model Override

Override the configured model for a single operation:

bashcopy
elith --model claude explain src/
elith --model lmstudio refactor src/cache/

// Configuration Locations

Elith checks for configuration in this order:

  1. ~/.elith/config.toml (user config)
  2. Environment variables (.env or exported)
  3. Default values (LM Studio on localhost:1234)

// Verify Configuration

Check which models are configured:

bashcopy
elith models

Output:

Provider    Status
claude      ✓ Configured
lmstudio    ✓ Configured
openrouter  Not configured

// Advanced: Manual Configuration

If you prefer to edit the config file directly:

# Edit config
nano ~/.elith/config.toml

# Restart backend to apply changes
elith service restart

// Troubleshooting

Configuration Not Found

# Recreate config
elith init

Invalid API Key

# Reconfigure provider
elith init

Backend Not Using New Config

# Restart backend
elith service restart

// Next Steps