swobuforge

Swobu

Name: Swobu
Author: swobuforge

DevOps community

SOURCE https://github.com/swobuforge/swobu

Description

Provider router and compatibility proxy for Claude Code, Codex and AI coding agents. Switch and fail over across OpenAI, Anthropic, OpenRouter, Ollama, Bedrock, Azure and more without breaking tools, reasoning or sessions.

Installation

This entry records only its repository, not the path inside it, so there is no exact command to give. Open the source below and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

README

Swobu

**English** · [简体中文](README.zh-CN.md) · [日本語](README.ja.md) · [Português (Brasil)](README.pt-BR.md) · [Bahasa Indonesia](README.id.md) · [한국어](README.ko.md) · [Русский](README.ru.md) · [Español](README.es.md) · [Українська](README.uk.md)

**One endpoint for your AI agents. Any LLM capacity underneath.**

Make AI capacity routable. Your agent asks for a model. Swobu turns that model name into a route across providers, accounts, regions, and local servers — with balancing, failover, reasoning translation, and semantic protocol compatibility underneath.

An AI agent using a Swobu route with balancing and failover

[Documentation](https://swobu.com/docs/) · [Quickstart](https://swobu.com/docs/start/first-route/) · [Releases](https://github.com/swobuforge/swobu/releases)

Agents and clients supported by Swobu Providers supported by Swobu

Your agent chooses a model. Swobu chooses where it runs.

A Swobu **route looks like a model** to your agent.

Behind that name can be one endpoint, the same model available from several places, or a cross-provider pool.

claude-opus-5
    │
    ├─ Anthropic / claude-opus-5
    ├─ AWS Bedrock / account A / claude-opus-5
    └─ AWS Bedrock / account B / claude-opus-5

Keep using `claude-opus-5`. Swobu can balance capacity and fail over underneath it.

Or make the model name describe a job:

codex-auto-review
    │
    ├─ Deepseek / Deepseek V4 Flash
    ├─ Google / Gemini 3.7 Flash
    └─ another review model

Or build a pool that deliberately crosses models and providers:

free
    │
    ├─ Cerebras / Gemma 4 31B
    ├─ Groq / gpt-oss-20b
    ├─ LLM7 / default
    ├─ OpenRouter / free
    ├─ Mistral / Ministral 3B
    ├─ NVIDIA NIM / Nemotron Mini 4B
    └─ O