Laroute

Use one base URL and key format across all connected providers.

Connect OpenRouter, OpenAI-compatible services, and private vLLM servers.

Create virtual keys with expiry dates and per-period budgets.

Shared app accounts with their own keys, budgets, and request history.

See model, key, result, token use, cost, and response time per request.

Detect tools like Cursor and Cline, and rank apps by tokens and spend.

Refresh provider model lists hourly with admin approval.

Get notified when providers are down or error rates climb.

Runs as one Go application with SQLite and an included Docker image.

Chat directly with models, with persistent conversations and live budget insight.

Personalize chats with reusable skills and connect external tools.

Manage keys, models, and users with a guided first-run setup.

Separate virtual keys for people and apps.

Provider keys stay hidden from users.

Add expiry dates to keys.

Set daily, weekly, or monthly spending limits.

Stop new requests when the budget is used up.

Users manage their own keys through an optional API.

Team accounts with their own keys and budgets.

Keep a request history for each app.

Members see the apps they belong to.

Chat with models directly, alongside the API gateway.

Pick models and adjust reasoning effort right from the composer.

See context usage and budget status while chatting.

View pricing and context limits before choosing a model.

Replies keep streaming through refresh and resume when you return.

Compact long conversations to continue past the context limit.

Regenerate answers and keep alternate versions.

See when the assistant uses a tool and what it returned.

Render model tables cleanly and attach videos to conversations.