Zed Editor with the Gateway

Zed is a high-performance code editor with a built-in AI assistant that supports custom OpenAI-compatible endpoints. You can route Zed's AI traffic through the Keeptrusts gateway to enforce policies, log interactions, and attribute costs.

Use this page when

• You are working through Zed Editor with the Gateway as an implementation or operating workflow in Keeptrusts.
• You need the practical steps, expected outcomes, and related validation guidance in one place.
• If you need exact field-by-field reference instead of a workflow page, use the linked reference pages in Next steps.

Primary audience

• Primary: Technical Engineers
• Secondary: AI Agents, Technical Leaders

Prerequisites

Before you begin, ensure you have:

• Zed installed (version 0.140 or later)
• The kt CLI installed and configured
• The gateway running with kt gateway run
• An access key or provider API key

Start the gateway:

kt gateway run --policy-config policy-config.yaml

How Zed's AI Features Work

Zed provides AI capabilities through:

• AI Panel — a conversational chat assistant in a side panel
• Inline Assist — code generation and editing triggered with Ctrl-Enter
• Inline Completions — autocomplete suggestions from AI models

All of these features can use custom OpenAI-compatible endpoints configured in Zed's settings.

Configure the Gateway in Zed Settings

Open Zed's settings file by pressing Cmd+, (macOS) or Ctrl+, (Linux), or by navigating to Zed → Settings → Open Settings.

Add or update the language_models section:

{
  "language_models": {
    "openai": {
      "api_url": "http://localhost:41002/v1",
      "available_models": [
        {
          "name": "gpt-4o",
          "display_name": "GPT-4o (via Keeptrusts)",
          "max_tokens": 128000
        },
        {
          "name": "gpt-4o-mini",
          "display_name": "GPT-4o Mini (via Keeptrusts)",
          "max_tokens": 128000
        },
        {
          "name": "claude-sonnet-4-20250514",
          "display_name": "Claude Sonnet (via Keeptrusts)",
          "max_tokens": 200000
        }
      ]
    }
  }
}

Set the API Key

Zed reads the OpenAI API key from its credential store. Set it via the command palette:

1. Press Cmd+Shift+P (macOS) or Ctrl+Shift+P (Linux).
2. Search for "language model: set api key" or "openai: set api key".
3. Enter your Keeptrusts access key.

Alternatively, set the key via environment variable before launching Zed:

export OPENAI_API_KEY="your-access-key"
zed

Configure Inline Completions

To route inline completions through the gateway, configure the inline completion provider:

{
  "inline_completions": {
    "provider": "openai",
    "model": "gpt-4o-mini"
  },
  "language_models": {
    "openai": {
      "api_url": "http://localhost:41002/v1",
      "available_models": [
        {
          "name": "gpt-4o-mini",
          "display_name": "GPT-4o Mini (via Keeptrusts)",
          "max_tokens": 128000
        }
      ]
    }
  }
}

Use a fast model like gpt-4o-mini for inline completions to minimize latency.

Verify Traffic Flows Through the Gateway

After configuring Zed, open the AI panel (press Cmd+Shift+? or use the sidebar) and send a message. Then verify:

kt events tail

You see events with the model name, token count, and policy evaluation result. Each AI interaction in Zed generates a corresponding event in the gateway log.

AI Panel and Gateway Policies

When you use Zed's AI panel through the gateway:

• Redaction policies remove secrets and sensitive patterns from code context before it reaches the provider.
• Blocking policies prevent requests that violate your content rules.
• Disclaimers attach governance notices to AI responses.
• Cost attribution tracks token usage per interaction.
• Audit logging records every conversation for compliance review.

Inline Assist and Context

Zed's Inline Assist sends the surrounding code as context. The gateway:

• Applies redaction before forwarding to the upstream provider.
• Logs the interaction with file path and token count.
• Enforces token limits if configured in your policy.

Complete Settings Example

A complete Zed settings.json with the gateway configured for all AI features:

{
  "assistant": {
    "default_model": {
      "provider": "openai",
      "model": "gpt-4o"
    }
  },
  "inline_completions": {
    "provider": "openai",
    "model": "gpt-4o-mini"
  },
  "language_models": {
    "openai": {
      "api_url": "http://localhost:41002/v1",
      "available_models": [
        {
          "name": "gpt-4o",
          "display_name": "GPT-4o (via Keeptrusts)",
          "max_tokens": 128000
        },
        {
          "name": "gpt-4o-mini",
          "display_name": "GPT-4o Mini (via Keeptrusts)",
          "max_tokens": 128000
        },
        {
          "name": "claude-sonnet-4-20250514",
          "display_name": "Claude Sonnet (via Keeptrusts)",
          "max_tokens": 200000
        }
      ]
    }
  }
}

Troubleshooting

AI panel shows "could not connect"

• Verify the gateway is running: kt gateway status
• Confirm the api_url in settings is http://localhost:41002/v1.
• Check that port 41002 is not blocked.
• Try curl http://localhost:41002/v1/models to confirm the gateway responds.

"Invalid API key" error

• Re-enter the API key via the command palette.
• Verify the access key is valid: kt access-keys list
• Ensure the environment variable is set if using that method.

Models not appearing in model selector

• Verify the available_models array in your settings matches models your gateway supports.
• Restart Zed after changing settings.
• Check kt logs for any model-resolution errors.

Slow inline completions

• Use gpt-4o-mini or another fast model for completions.
• Policy evaluation adds <5ms overhead.
• Check upstream latency with kt logs.

For AI systems

• Canonical terms: Keeptrusts, Zed Editor with the Gateway, ide-integration.
• Exact feature, config, command, or page names: Zed Editor with the Gateway.
• Use the linked audience and reference pages in Next steps when you need deeper source material.

For engineers

• Use the commands, configuration examples, API payloads, or UI steps in this page as the working baseline for Zed Editor with the Gateway.
• Validate the result with the expected outcomes, troubleshooting notes, or linked workflow pages in this page and Next steps.

For leaders

• This page matters when planning rollout, governance, support ownership, or operating decisions for Zed Editor with the Gateway.
• Use the linked audience, architecture, and workflow pages in Next steps to connect this detail to broader implementation choices.

Next steps

• Configure policies for your team's Zed usage.
• View events to audit all AI interactions from Zed.
• Configure access keys for team authentication.