> ## Documentation Index
> Fetch the complete documentation index at: https://docs.scitix.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Zed

> Learn how to configure Zed with ScitiX Model Inference for AI-assisted coding through an OpenAI-compatible provider.

[Zed](https://zed.dev) is a high-performance multiplayer code editor with a built-in AI assistant. It supports OpenAI-compatible providers, so you can point its assistant at the ScitiX Model Inference API for chat and edits inside the editor.

Supported models include `glm-5.2` and `DeepSeek-V4-Flash` — see the [Models](https://console.scitix.ai/model-inference/models) page for the full list and pricing.

## Configure

Create an API key on the ScitiX [API Keys](https://console.scitix.ai/model-inference/api_keys) page first.

### 1. Add the provider

Open Zed's `settings.json` (command palette → **zed: open settings**) and add an OpenAI-compatible provider under `language_models`:

```json theme={null}
{
  "language_models": {
    "openai_compatible": {
      "scitix": {
        "api_url": "https://api.scitix.ai/model-api/v1",
        "available_models": [
          {
            "name": "glm-5.2",
            "display_name": "ScitiX glm-5.2",
            "max_tokens": 131072
          }
        ]
      }
    }
  }
}
```

| Field                           | Description                                                                                                                              |
| ------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------- |
| `openai_compatible.<id>`        | Provider ID, e.g. `scitix`. The API-key environment variable name is derived from it                                                     |
| `api_url`                       | ScitiX OpenAI-compatible endpoint: `https://api.scitix.ai/model-api/v1`                                                                  |
| `available_models[].name`       | Model ID as served by ScitiX, e.g. `glm-5.2`. Choose a model with native tool calling if you rely on the assistant's tool/edit features. |
| `available_models[].max_tokens` | Context window for the model                                                                                                             |

Models default to `tools: true`, so tool calls work without extra capability flags. Set `max_tokens` to the model's real context window.

### 2. Provide the API key

**Do not put the API key in `settings.json`.** Zed accepts it two ways:

* **Settings UI** — open **Settings → User → AI → General → LLM Providers**, find **scitix** in the list, and click **Configure** to enter the key. Zed stores it in the system keychain.
* **Environment variable** — Zed derives the name from the provider ID in upper snake case plus `_API_KEY`. For the `scitix` provider above, set `SCITIX_API_KEY`, then launch Zed from a shell where that variable is exported.

The same provider can also be added from the UI (**Settings → AI → LLM Providers → Add Provider → OpenAI**), which stores the key in the system keychain:

<img src="https://mintcdn.com/scitix-adfc65c2/FurCT4ZVEUYC3l_M/model-inference/media/zed-provider-config.png?fit=max&auto=format&n=FurCT4ZVEUYC3l_M&q=85&s=c74153b001299bb9154beb830e1a8a4d" alt="Zed — OpenAI-Compatible provider pointed at ScitiX" width="1800" height="1500" data-path="model-inference/media/zed-provider-config.png" />

## Verify

Open the Agent Panel, select the ScitiX model (shown as **ScitiX glm-5.2**) from the model picker, and send a message — a normal reply confirms the connection.

<img src="https://mintcdn.com/scitix-adfc65c2/FurCT4ZVEUYC3l_M/model-inference/media/zed-agent-reply.png?fit=max&auto=format&n=FurCT4ZVEUYC3l_M&q=85&s=9fa67855bd466924e7c4d06390c6a466" alt="Zed Agent Panel — coding with ScitiX glm-5.2" width="3024" height="1694" data-path="model-inference/media/zed-agent-reply.png" />

If the model is not listed, confirm the `available_models` entry is present and `api_url` ends with `/v1`.

Verified on Zed 1.14.2 (macOS); the AI settings layout may shift between versions.