Zed Editor

This screenshot will be refreshed with kineticRouter branding during the integration documentation pass.

Why Zed?

Zed offers one of the most complete AI coding experiences available:

Native agent support — Built-in Claude Code and Codex CLI as external agents
Custom LLM Provider — Connect any model via OpenAI compatible protocol
High performance — Written in Rust, fast startup and low resource usage

Paired with kineticRouter, one API Key gives you access to leading global models in Zed.

Quick Start

1. Get an API Key

Go to the kineticRouter Console to create an API Key.

2. Add Provider

Agent Panel (Recommended)

Add visually through the Agent Panel without editing JSON.

Press ⌘ + Shift + A to open the Agent Panel
Click + Add Provider in the LLM Providers area
Fill in the configuration as shown:

This screenshot will be refreshed with kineticRouter branding during the integration documentation pass.

Label	Field	Value
①	Provider Name	`kineticRouter`
②	API URL	`https://api.kineticrouter.com/v1`
③	API Key	Your kineticRouter API Key
④	Model Name	`openai/gpt-5.4-mini`
⑤	Max Completion Tokens	`512000`
⑥	Capabilities	Check the capabilities supported by the model

Click + Add Model to add more models.

The API Key is securely stored in the system keychain (macOS Keychain / Linux Secret Service) and is never written in plaintext to config files.

Getting Started

Zed’s Agent Panel has two modes:

Mode	Description
Zed Agent	Built-in AI assistant using your configured LLM Provider (e.g. kineticRouter)
External Agents	External agents such as Claude Code and Codex CLI, run independently

Start a conversation with Zed Agent:

This screenshot will be refreshed with kineticRouter branding during the integration documentation pass.

Step	Action
①	Click + → Select Zed Agent (`⌘ + N`)
②	External Agents area shows connected external agents
③	Bottom-right model selector → kineticRouter group → Select model

This screenshot will be refreshed with kineticRouter branding during the integration documentation pass.

Recommended Models

For recommended models, see the Model Marketplace .

Adding More Models

Each model requires the following parameters:

Field	Required	Description
`name`	Yes	Model ID, e.g. `openai/gpt-5.4-mini`
`display_name`	No	UI display name
`max_tokens`	Yes	Context window size
`max_output_tokens`	No	Maximum output token count
`capabilities.tools`	No	Whether Function Calling is supported
`capabilities.images`	No	Whether image input is supported

Pro Tip

Manually filling parameters too slow? Click Copy Page in the top right of this page (or send this page URL directly), along with https://api.kineticrouter.com/v1/models, to an AI and have it auto-generate the complete settings.json configuration.

Go to settings → Click Edit in settings.json in the top right to paste:

This screenshot will be refreshed with kineticRouter branding during the integration documentation pass.

Troubleshooting

kineticRouter not showing in model list

Verify settings.json format is correct, then restart Zed.

Authentication error

Command Palette ⌘ + Shift + P → Search language model: reset credentials → Re-enter your API Key.

Model doesn’t support tool calling

Set capabilities.tools to false.