# ModelMux Configuration Example
# Copy to /etc/modelmux/config.toml (system-wide) or ~/.config/modelmux/config.toml (user)
# Then run: sudo systemctl enable --now modelmux (system) or systemctl --user enable --now modelmux (user)
[server]
# HTTP server port (default: 3000)
port = 3000
# Log level: trace, debug, info, warn, error (default: info)
log_level = "info"
# Enable automatic retries on transient failures (default: true)
enable_retries = true
# Maximum number of retry attempts (default: 3)
max_retry_attempts = 3
[auth]
# Option 1: Path to Google Cloud service account JSON file
# service_account_file = "/etc/modelmux/service-account.json"
# Option 2: Inline service account JSON (recommended for containers/systemd)
# Get your service account JSON from Google Cloud Console:
# 1. Go to https://console.cloud.google.com/
# 2. IAM & Admin → Service Accounts
# 3. Create/select account with "Vertex AI User" role
# 4. Download JSON key and paste here as a single line
service_account_json = "{\"type\": \"service_account\", \"project_id\": \"your-project-id\", \"private_key_id\": \"key-id\", \"private_key\": \"-----BEGIN PRIVATE KEY-----\\n...\\n-----END PRIVATE KEY-----\\n\", \"client_email\": \"your-service-account@your-project.iam.gserviceaccount.com\", \"client_id\": \"123456789\", \"auth_uri\": \"https://accounts.google.com/o/oauth2/auth\", \"token_uri\": \"https://oauth2.googleapis.com/token\", \"auth_provider_x509_cert_url\": \"https://www.googleapis.com/oauth2/v1/certs\", \"client_x509_cert_url\": \"https://www.googleapis.com/robot/v1/metadata/x509/your-service-account%40your-project.iam.gserviceaccount.com\", \"universe_domain\": \"googleapis.com\"}"
[streaming]
# Streaming mode: auto (detect client), never, standard, buffered, always
# auto = automatically detect client capabilities (recommended)
mode = "auto"
# Buffer size for streaming responses (default: 8192 bytes)
buffer_size = 8192
# Timeout for streaming chunks in milliseconds (default: 1000)
chunk_timeout_ms = 1000
[vertex]
# Google Cloud project ID (required)
project = "your-gcp-project-id"
# Vertex AI region (required)
# Common regions: us-central1, us-east1, europe-west1, europe-west4, asia-northeast1
region = "us-central1"
# Vertex AI location (usually same as region)
location = "us-central1"
# Model publisher (default: anthropic)
# Available: anthropic, google, meta
publisher = "anthropic"
# Model ID (required)
# Anthropic Claude models:
# claude-3-5-sonnet-20241022 # Latest Claude 3.5 Sonnet
# claude-3-5-haiku-20241022 # Latest Claude 3.5 Haiku
# claude-3-opus-20240229 # Claude 3 Opus
# claude-3-sonnet-20240229 # Claude 3 Sonnet
# claude-3-haiku-20240307 # Claude 3 Haiku
model = "claude-3-5-sonnet-20241022"
# Alternative: Full URL override (replaces all above vertex settings)
# url = "https://us-central1-aiplatform.googleapis.com/v1/projects/your-project/locations/us-central1/publishers/anthropic/models/claude-3-5-sonnet-20241022"
# ==============================================================================
# SETUP INSTRUCTIONS
# ==============================================================================
#
# 1. GOOGLE CLOUD SETUP:
# - Create a GCP project with billing enabled
# - Enable the Vertex AI API
# - Create a service account with "Vertex AI User" role
# - Download the service account JSON key
#
# 2. CONFIGURATION:
# - Copy this file to /etc/modelmux/config.toml
# - Edit the [vertex] section with your project details
# - Edit the [auth] section with your service account JSON
# - Adjust [server] and [streaming] settings as needed
#
# 3. INSTALLATION (choose one):
# a) System-wide (recommended):
# sudo cp config.toml /etc/modelmux/
# sudo chmod 644 /etc/modelmux/config.toml
# sudo systemctl enable --now modelmux
#
# b) User-specific:
# mkdir -p ~/.config/modelmux
# cp config.toml ~/.config/modelmux/
# systemctl --user enable --now modelmux
#
# 4. VERIFICATION:
# modelmux doctor # Check configuration
# systemctl status modelmux # Check service status
# curl -X POST http://localhost:3000/v1/chat/completions \
# -H "Content-Type: application/json" \
# -H "Authorization: Bearer dummy" \
# -d '{"model":"claude-3-5-sonnet-20241022","messages":[{"role":"user","content":"Hello!"}],"max_tokens":50}'
#
# 5. LOGS:
# journalctl -u modelmux -f # Follow system logs
# journalctl --user -u modelmux -f # Follow user logs
#
# For more information: https://github.com/yarenty/modelmux