Integration guide · Code editor

Zed on EU-hosted open models

Add Frontière AI as an OpenAI-compatible provider in Zed's settings.json and drive the agent panel with EU-hosted open models. language_models.openai_compatible in ~/.config/zed/settings.json; the key in the keychain or FRONTIERE_API_KEY.

Settings from the official documentation, checked on 12 October 2026. Not run end to end by us.

API format

OpenAI-compatible

Base URL

https://getfrontiereai.eu/api/v1

Where it goes

language_models.openai_compatible in ~/.config/zed/settings.json; the key in the keychain or FRONTIERE_API_KEY.

Your prompts

Straight from your machine to us

Setup

// ~/.config/zed/settings.json
{
  "language_models": {
    "openai_compatible": {
      "frontiere": {
        "api_url": "https://getfrontiereai.eu/api/v1",
        "available_models": [
          {
            "name": "glm-5.2",
            "display_name": "GLM 5.2 (Frontière AI)",
            "max_tokens": 256000,
            "capabilities": { "tools": true, "images": false }
          }
        ]
      }
    }
  }
}

Which model to pick

24 live models have tool calling measured by a real call through the gateway. Models labeled EU sovereign run at European providers; fast-access models are served through US-controlled infrastructure. Prices are per 100,000 tokens, margin and VAT included.

Good to know

• Edit predictions (Tab) call a completions endpoint, not chat completions: they will not work with Frontière AI.
• External agents run inside Zed (Claude Code, for instance) ignore these settings: configure the agent itself.
• For a reasoning model, Zed sends max_completion_tokens by default; set capabilities.max_tokens_parameter if needed.

Where your prompts go

Zed's privacy page says that with your own API key the configured provider handles requests, and providers are initialised in the local app. Only Zed-hosted models go through Zed.

On our side, a model labeled EU sovereign runs at a European provider, and the gateway itself runs in Paris on Vercel, a US company — our privacy page says so. Billing keeps the model, the token counts and the amount, never the prompt or the reply.

Sovereignty →Privacy →

Sources

Official documentation, checked on 12 October 2026:

• Zed — Use API access
• Zed — Use a gateway

FAQ