Loco is a web and API framework running on Rust.
This is the SaaS starter which includes a User model and authentication based on JWT.
It also include configuration sections that help you pick either a frontend or a server-side template set up for your fullstack server.
cargo loco start$ cargo loco start
Finished dev [unoptimized + debuginfo] target(s) in 21.63s
Running `target/debug/myapp start`
:
:
:
controller/app_routes.rs:203: [Middleware] Adding log trace id
▄ ▀
▀ ▄
▄ ▀ ▄ ▄ ▄▀
▄ ▀▄▄
▄ ▀ ▀ ▀▄▀█▄
▀█▄
▄▄▄▄▄▄▄ ▄▄▄▄▄▄▄▄▄ ▄▄▄▄▄▄▄▄▄▄▄ ▄▄▄▄▄▄▄▄▄ ▀▀█
██████ █████ ███ █████ ███ █████ ███ ▀█
██████ █████ ███ █████ ▀▀▀ █████ ███ ▄█▄
██████ █████ ███ █████ █████ ███ ████▄
██████ █████ ███ █████ ▄▄▄ █████ ███ █████
██████ █████ ███ ████ ███ █████ ███ ████▀
▀▀▀██▄ ▀▀▀▀▀▀▀▀▀▀ ▀▀▀▀▀▀▀▀▀▀ ▀▀▀▀▀▀▀▀▀▀ ██▀
▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀
https://loco.rs
environment: development
database: automigrate
logger: debug
compilation: debug
modes: server
listening on http://localhost:5150You can check your configuration to pick either frontend setup or server-side rendered template, and activate the relevant configuration sections.
AI providers are configured once per app instance under settings.ai; browser clients and individual users cannot supply provider endpoints.
settings:
ai:
default_provider: ollama
providers:
- id: ollama
name: Local Ollama
kind: ollama
base_url: http://host.docker.internal:11434
default_model: qwen3.8:27bProjects store the provider id, an optional model override, and their own system prompt. Multiple Ollama instances can already be configured by adding entries with distinct IDs. Additional provider kinds are added as backend adapters without changing projects or the frontend response format.
Check out a quick tour or the complete guide.