Best open-source AI repos for LLM gateways and model routers (2026)

What is the best open-source LLM gateway or router?

The best open-source AI repos for LLM gateways and model routers right now are Litellm, CC Switch and Experiential. LiteLLM is the standard gateway pick: one OpenAI-format API for 100+ providers with cost tracking; CC Switch manages model and provider config for Claude Code, Codex and other coding tools from one desktop app.

Comparison

Repo	What it does	Stars	Start here
Litellm BerriAI/litellm · Python	LiteLLM is an open source AI Gateway and Python SDK that lets you call 100+ LLM providers through one OpenAI-format interface, with cost tracking, guardrails, and load balancing.	61,803	Call 100+ LLM providers through one OpenAI-format API Intermediate, 30 minutes
CC Switch farion1231/cc-switch · Rust	CC Switch is a desktop app for Windows, macOS and Linux that manages the config of Claude Code, Claude Desktop, Codex, Gemini CLI, OpenCode, OpenClaw, Hermes Agent and other coding CLIs.	148,400	Switch Claude Code or Codex providers in one click Beginner, 10 minutes
Experiential experientiallabs/experiential · Python	Experiential is an open source gateway and router for agent workflows.	13,003	Run one local gateway in front of all your models Beginner, 15 minutes
Free Claude Code Alishahryar1/free-claude-code · Python	FCC is an open-source Python proxy that lets coding agents such as Claude Code and Codex use free, paid, and local models from many providers through one Admin UI.	62,032	Run Claude Code on free models through FCC Intermediate, 30 minutes

When to pick which

Litellm BerriAI/litellm

LiteLLM is an open source AI Gateway and Python SDK that lets you call 100+ LLM providers through one OpenAI-format interface, with cost tracking, guardrails, and load balancing.

Best for: Developers and platform teams who call several LLM providers and want one API format, or who need a shared gateway for a team.

Needs: Python and the uv package manager installed; API keys for the LLM providers you plan to call, such as OpenAI or Anthropic; For the gateway use cases: a machine to run the proxy server on and a LiteLLM master key or virtual key for clients

Start here: Call 100+ LLM providers through one OpenAI-format API Intermediate, 30 minutes

Install the SDK with uv add litellm
Set the provider keys as environment variables, such as OPENAI_API_KEY and ANTHROPIC_API_KEY
Call completion from Litellm with a model such as openai/gpt-4o
Change the model string to anthropic/claude-sonnet-4-20250514 and run the same call again

All 5 use cases GitHub

CC Switch farion1231/cc-switch

CC Switch is a desktop app for Windows, macOS and Linux that manages the config of Claude Code, Claude Desktop, Codex, Gemini CLI, OpenCode, OpenClaw, Hermes Agent and other coding CLIs. It switches API providers in one click from 90+ presets, can route requests locally with format conversion and failover, and manages MCP servers, Skills and Prompts in one place, so you stop hand-editing JSON, TOML and YAML files. The only official site is ccswitch.io.

Best for: Developers who use more than one coding agent or more than one model provider and keep breaking config files by hand.

Needs: Windows 10+, macOS 12+ or Linux with glibc 2.35+; API keys or plans for the providers you add

Start here: Switch Claude Code or Codex providers in one click Beginner, 10 minutes

Download CC Switch from the GitHub releases page (macOS users can use Homebrew)
Open the page for Claude Code, Codex or another tool
Add a provider from the presets and enter its key
Click to switch; only the endpoint, key and model change, and the original config is backed up first

All 3 use cases GitHub

Experiential experientiallabs/experiential

Experiential is an open source gateway and router for agent workflows. It puts hosted, BYOK and local models behind one OpenAI-compatible API, lets you control which users and agents can use which models and how much they can spend, and turns production traffic into a custom router or model optimized for quality, speed and cost.

Best for: Developers and teams running coding agents or LLM apps who want one endpoint for many models, with spending limits and a path to cheaper routing.

Needs: Python with pip to install the package; Provider keys for the models you want to route to (hosted, BYOK or local); For Capture: macOS, Python 3.13 or newer, and approval for its extension and certificate; Optional: an account on the hosted platform, and Tinker for fine-tuning an open source model

Start here: Run one local gateway in front of all your models Beginner, 15 minutes, works with claude-code, codex, cursor

Run pip install experiential
Run exp to start a local OpenAI-compatible gateway; the setup wizard picks provider, model and reasoning effort, then shows defaults for the public alias, identity and the $50.00 command budget before printing a one-time key
Choose a public alias such as opus-5 and capture the issued key
Set the key with export EXP_GATEWAY_KEY=..., then send the curl request from the README to http://127.0.0.1:8000/v1/chat/completions with your alias as the model and the key in the Authorization header

All 3 use cases GitHub

Free Claude Code Alishahryar1/free-claude-code

FCC is an open-source Python proxy that lets coding agents such as Claude Code and Codex use free, paid, and local models from many providers through one Admin UI.

Best for: Developers and power users who want to run terminal, IDE, or browser coding agents on low-cost or free model providers.

Needs: A macOS, Linux, or Windows computer with a terminal; An API key from at least one provider in the README catalog, such as NVIDIA NIM; Permission to run the install script from the README

Start here: Run Claude Code on free models through FCC Intermediate, 30 minutes, works with claude-code

Run curl -fsSL "https://raw.githubusercontent.com/Alishahryar1/free-claude-code/main/scripts/install.sh" | sh on macOS or Linux, or the PowerShell command from the README on Windows, and choose Claude Code when prompted.
Run fcc-server on Linux, or open Free Claude Code from the desktop or Start menu on Windows or macOS.
Open the Admin UI URL from the server log, open NVIDIA NIM under Providers, click Add key, paste your key, and click Save.
Leave MODEL on the default or search the dropdown, then click Apply.

All 4 use cases GitHub

Use cases for LLM gateways and model routers

Call 100+ LLM providers through one OpenAI-format API Litellm
Run a shared self-hosted AI Gateway for your team Litellm
Give LLM calls access to MCP server tools Litellm
Switch Claude Code or Codex providers in one click CC Switch
Add MCP servers and Skills once, sync them to every agent CC Switch
See token usage, spend and quotas across your coding agents CC Switch
Run one local gateway in front of all your models Experiential
Repoint Claude Code, Cursor and Codex at one gateway Experiential
Turn your agent traces into a cheaper router Experiential
Run Claude Code on free models through FCC Free Claude Code
Keep a coding session running during provider outages Free Claude Code
Run a Codex session on a project folder in the browser Free Claude Code

Questions

What is the best open-source LLM gateway or router?

Which repo for LLM gateways and model routers is easiest to start with?

CC Switch: its starter use case "Switch Claude Code or Codex providers in one click" is rated Beginner and takes about 10 minutes.

How do I get started with Litellm?

Install the SDK with uv add litellm; Set the provider keys as environment variables, such as OPENAI_API_KEY and ANTHROPIC_API_KEY; Call completion from Litellm with a model such as openai/gpt-4o. Full steps and a copy-paste agent prompt: https://dedupedai.com/use-cases/berriai-litellm/call-100-llms-with-one-openai-format-api/

Are these LLM gateways and model routers repos free?

They are open-source projects on GitHub (check each license; some, like fair-code projects, limit commercial use). Some also sell a hosted version or need a paid model API key; the requirements for each repo are listed on this page.

How is this list of LLM gateways and model routers repos chosen?

Deduped tracks the hottest AI repos on GitHub and writes step-by-step use cases from each README. For this page we picked the repos that do this job, ordered by fit and then GitHub stars. Repos we only inferred from a short README, or that we do not index, are left out. The page is rebuilt with the site; stars are as of the last build in 2026.

What is the best open-source LLM gateway or router?

Comparison

When to pick which

Litellm BerriAI/litellm

CC Switch farion1231/cc-switch

Experiential experientiallabs/experiential

Free Claude Code Alishahryar1/free-claude-code

Use cases for LLM gateways and model routers

Questions

More answers