Name: 'Z-AI — One key for every LLM'
Author: 'ZyoraLabs'

One key. Every LLM. Inside your editor.

Chat, code actions and inline completions powered by the Z-AI gateway — OpenAI, Anthropic, Gemini, Groq, Mistral, Together, Fireworks and Nexula behind one API key.

Features

Sidebar chat with streaming responses, multi-turn history and "attach selection" support
Code actions on any selection — Explain, Refactor, Generate tests, Add comments
Status bar model picker — switch between any model the gateway exposes
Inline completions (opt-in) — ghost-text suggestions as you type
Secure key storage via VS Code SecretStorage (your key never hits a settings file)

Quick start

Install: search "Z-AI" in the Extensions sidebar.
Get an API key from the Z-AI console or via the CLI:
```
npm install -g @zyoralabs/zai-cli
zai keys list
```
In VS Code: open the command palette → Z-AI: Sign in → paste your zk_live_… key.
Click the Z-AI icon in the activity bar to open chat, or right-click any selection for code actions.

Settings

Setting	Default	Notes
`zai.baseUrl`	`http://127.0.0.1:8011/v1`	Gateway base URL — set to your production endpoint
`zai.defaultModel`	`openai/gpt-5.4-mini`	Used for chat and code actions
`zai.completionModel`	`openai/gpt-5.4-mini`	Used for inline completions (pick something fast)
`zai.completions.enabled`	`false`	Toggle ghost-text completions
`zai.completions.debounceMs`	`350`	Wait before requesting a completion
`zai.completions.maxTokens`	`96`	Max tokens per completion
`zai.completions.contextLines`	`60`	Lines of context (before+after cursor)