⭐ Please give this project a star!

Router Models is 100% free. If it's useful to you, the best way to say thanks is a ⭐ Star on GitHub — it takes one second and helps other people find the project.

👉 Github — hit the ★ button!

🤔 What does it do?

VS Code Copilot Chat normally only offers Microsoft's own models. Router Models lets you add your own AI providers and use their models in the same chat — right from the regular model picker.

Why you'll love it:

• 🔑 Use several API keys together — got multiple (free) keys? Paste them all in. When one hits its limit, the next one takes over automatically. No more "rate limit" interruptions.
• ❄️ No manual waiting — if a key needs a break, the extension handles the timing and switches to another key for you.
• 🏠 Works with local AI too — use models running on your own computer (Ollama, LM Studio) — no key needed at all.
• 📊 See what's happening — a small indicator in the status bar shows your keys' status at a glance.
• 🔐 Your keys stay safe — they're stored in VS Code's secure storage, never in plain text files.
• 🧠 Works with thinking models and tool-using agents out of the box.

📦 How to Install

Option 1 — From the VS Code Marketplace (easiest)

1. Open VS Code
2. Click the Extensions icon on the left sidebar (or press Ctrl+Shift+X)
3. Search for Router Models
4. Click Install

Or open this link in your browser and click Install: https://marketplace.visualstudio.com/items?itemName=alireza-yaghouti.router-models

Option 2 — Download the file from GitHub

1. Go to the Releases page and download the file ending in .vsix
2. Open VS Code → click the Extensions icon (Ctrl+Shift+X)
3. Click the ⋯ menu (top of the Extensions panel) → Install from VSIX…
4. Choose the file you downloaded — done!

[!TIP] With Option 2, VS Code won't update the extension automatically. To get notified about new versions, click Watch → Custom → Releases on the GitHub page.

🚀 How to Use

1. Press Ctrl+Shift+P and type Router Models: Add Provider
2. Fill in the simple form:
   • Name — anything you like, e.g. My OpenAI
   • Base URL — the provider's address (see the table below)
   • API keys — one or more, each on a new line (skip this for local Ollama / LM Studio)
3. Open Copilot Chat, click the model picker at the top — your new models are there under Router Models. Pick one and chat! 🎉

You can also manage everything from the Router Models panel in the left activity bar — add, edit or remove providers with simple buttons.

🌐 Ready-to-copy Base URLs

Any other service that works with the "OpenAI format" will work too.

🏷️ Mark models as free

Every model in the Router Models panel has a small + free button. Click it to tag the model as free:

• the sidebar shows a green ✓ free badge next to the model,
• in the Copilot model picker the model is listed as Model name (free), so typing free in the picker's search box lists all your free models across every provider,
• models whose id or name already contains "free" (like OpenRouter's :free models) get the label automatically,
• the flag is kept when you refresh the provider's model list.

To remove the label, click the button again. The + Model form and the Router Models: Add Model Manually command ask about the free tag too.

🎁 Detect free models automatically

Instead of tagging models one by one, Router Models can tag them for you. It downloads a curated list of free models maintained by the extension author, matches every provider you added by the base URL's domain (so https://openrouter.ai/api/v1 in the list matches a provider you added as https://openrouter.ai/v1), and tags all the listed models as free automatically.

The list is a service shipped with the extension — there is nothing to host and no URL to configure. It is off by default; nothing is downloaded until you turn it on.

1. Turn it on. Click Enable in the sidebar footer, press the 🎁 button in the sidebar header, or run the Router Models: Detect Free Models command. You can also flip the routerModels.freeModelsEnabled setting.

2. That's it. The extension then:

• shows a footer line with the model count, provider count, how long ago the list was updated, and an Update link to download the latest list on demand,
• re-downloads the list automatically every routerModels.freeModelsRefreshHours hours (24 by default — you can change it),
• refreshes on startup too when the cached copy is older than the configured interval, so a machine that was off for days still catches up.

A model that gets dropped from the list stops being free again, and your own manual + free tags keep working alongside it.

For the curious, the list is a plain JSON file that maps a provider's domain to its free model ids (glob patterns supported, so meta-llama/* marks every model under that prefix as free):

{
  "version": 1,
  "updated": "2025-01-30T12:00:00Z",
  "providers": [
    {
      "baseUrl": "https://openrouter.ai/api/v1",
      "freeModels": [
        "deepseek/deepseek-chat-v3:free",
        "qwen/qwen-2.5-72b-instruct:free",
        "meta-llama/*"
      ]
    },
    {
      "baseUrl": "https://api.deepseek.com",
      "freeModels": ["deepseek-chat", "deepseek-reasoner"]
    }
  ]
}

📥 Import from a backup file (9router, omnirouter or any fork of 9router)

1. Run Router Models: Import Providers from JSON — or click the ⇩ button in the sidebar.
2. Pick your .json backup file.
3. That's it! All providers and API keys in the file are added automatically — keys you already have are skipped, so nothing gets duplicated. The model list fills in by itself.

📤 Export your setup (backup & move to another PC)

1. Run Router Models: Export Providers to JSON — or click the ⤒ button in the sidebar header.
2. Choose whether to include the API keys — with keys means the file restores your full setup anywhere (keep it private!), without keys is safe to share.
3. Save the file. Import it on any other machine with Router Models: Import Providers from JSON — you can then merge with the existing setup or replace it completely (full restore). Endpoints, icons, cooldowns, manual models, "free" tags and the include/exclude patterns all come along.

☁️ Keep work & home in sync automatically

Router Models works with VS Code Settings Sync (your GitHub or Microsoft account):

• Provider configs, model lists, manual models and "free" tags are synced between your machines automatically.
• API keys are not synced by default — they live in each machine's secure storage. To sync them too, enable routerModels.syncApiKeys in the settings on every machine (keys then travel inside the encrypted Settings Sync backup).
• Make sure Settings Sync itself is turned on (gear icon → Backing up and Syncing… → sign in with GitHub/Microsoft).
• After a sync, open the Router Models sidebar (or run Router Models: Reload from Settings Sync) to pick up the changes immediately.

❓ Common Questions

🤝 Contribute

Found a bug or have an idea? Open an issue — all kinds of contributions are welcome!

⭐ Support the Project

[!TIP] If Router Models saves you time or money, please give it a ⭐ Star!

Stars keep the project alive and help other developers discover it.

👉 github.com/web-elite/router-models

Other ways to help:

• 🐞 Report bugs
• 💡 Suggest ideas for new features
• 📣 Tell your friends and colleagues about it

📄 License

MIT © Web Elite