README

AI Provider for OpenCode Zen

One API key. 57 frontier AI models, inside WordPress. An independent OpenCode Zen provider for the WordPress AI Client — brings GPT 5.x, Claude, Gemini 3.x, Qwen, MiniMax, Kimi, Grok, and DeepSeek to every AI-enabled plugin on your site.

Not affiliated with, endorsed by, or sponsored by OpenCode Zen.

What it does

WordPress 7.0 ships a built-in AI Client SDK: plugins ask it to "generate text from this prompt" without caring which AI service answers. This plugin registers OpenCode Zen as one of those services — an AI model aggregator where a single API key unlocks frontier models from OpenAI, Anthropic, Google, and more.

Activate it, paste one key, and every AI-enabled plugin on your site can generate content through 57 models — no separate accounts, no per-provider billing, no code.

Features

One key, every model

57 models across GPT 5.x, Claude, Gemini 3.x, Qwen, MiniMax, Kimi, Grok, and DeepSeek — one OpenCode Zen key instead of an account per provider
Live model discovery from the OpenCode Zen API (cached for one hour), with a built-in 57-model fallback so generation keeps working if the API is briefly unreachable
A sensible default model (gpt-5.5) preselected out of the box

Native WordPress integration

Registers as a first-class provider for the WordPress 7.0+ AI Client — any AI-enabled plugin uses it with zero extra wiring
Enter your key on Settings → OpenCode Zen, the core Settings → Connectors screen, or via the OPENCODE_ZEN_API_KEY environment variable
Connection-status indicator on the settings page and the provider logo on the Connectors screen
Standard generation controls: temperature, max tokens, top P, and presence/frequency penalties

Built to stay out of the way

OpenAI-compatible under the hood — text generation, system instructions, stop sequences, and function declarations
Ships as both a WordPress plugin and a Composer package for standalone PHP use
No SDK bloat — WordPress core provides the AI Client at runtime, and the plugin excludes its own copy to stay collision-safe

Requirements

PHP 7.4+
WordPress 7.0+ (the AI Client SDK ships in core)
- On older WordPress, install the WordPress AI Client plugin separately

Installation

Method	How
WordPress.org	Search "AI Provider for OpenCode Zen" under Plugins → Add New, then install and activate
Upload	Download the zip and upload it via Plugins → Add New → Upload Plugin
Composer	`composer require mralaminahamed/ai-provider-for-opencode-zen`

Quick start

Activate the plugin.
Open Settings → OpenCode Zen (or the core Settings → Connectors screen) and paste your API key. Get one at opencode.ai/zen/settings/api-keys.
Any AI-enabled plugin can now generate text through the opencode-zen provider:

use WordPress\AiClient\AiClient;

echo AiClient::prompt('Explain quantum computing')
    ->usingProvider('opencode-zen')
    ->generateTextResult()
    ->toText();

See docs/USAGE.md for the standalone-Composer setup and per-request options.

Configuration

Configure defaults on Settings → OpenCode Zen (option key opencode_zen_settings):

Setting	Range	Default
Default Model	any available model	`gpt-5.5`
Temperature	0.0–2.0	0.7
Max Tokens	1–200,000	4096
Top P	0.0–1.0	1.0
Presence Penalty	-2.0–2.0	0.0
Frequency Penalty	-2.0–2.0	0.0

The OPENCODE_ZEN_API_KEY environment variable takes priority over the stored key. Full resolution order is in docs/ARCHITECTURE.md.

Supported models

When a key is configured, the live model list comes straight from the OpenCode Zen API, so you always see the latest models. If the API is unavailable, a 57-model fallback keeps everything working:

GPT 5.x — Sol / Terra / Luna, plus Pro, Mini, Nano, and Codex variants
Claude — Fable 5, Opus 4.x, Sonnet 5 / 4.x, Haiku 4.5
Gemini 3.x — Flash, Flash Lite, and Pro
More — Grok, Qwen, DeepSeek, MiniMax, GLM, Kimi, and free-tier models

The complete, ID-by-ID catalogue is in docs/MODELS.md.

How it works

WordPress 7.0's AI Client is a provider-agnostic layer. Install one or more provider plugins — like this one — and any plugin built on the AI Client can generate content through whichever provider is active, switchable from the Connectors screen with no per-plugin reconfiguration. This plugin handles the OpenCode Zen side: credential resolution, live model discovery, and the OpenAI-compatible API calls. Details in docs/ARCHITECTURE.md.

Documentation

Doc	What's in it
Usage	WordPress + standalone Composer examples, API key resolution, generation options, credential filters
Models	How model discovery and caching work, and the full 57-model fallback catalogue
Architecture	Provider identity, file layout, request flow, runtime SDK dependency
Development	Build/test/lint/analysis scripts, the AI Client `replace` guard and dev-only stub install, CI and release

Release history: CHANGELOG.md.

Development

composer install
composer test        # PHPUnit
composer phpcs       # WordPress Coding Standards
composer phpstan     # static analysis
composer release     # build the distributable zip

Full workflow — including the AI Client replace guard and the dev-only stub install that keeps PHPStan and PHPUnit green — is in docs/DEVELOPMENT.md.

Contributing

Issues and pull requests are welcome:

Open a new issue to report a bug or request a feature
Submit a pull request — see docs/DEVELOPMENT.md for the local build and test workflow

Support

If this plugin is useful, star the repo and leave a review on WordPress.org. Questions and bug reports go in the issue tracker.

mralaminahamed / ai-provider-for-opencode-zen

Maintainers

Package info

Fund package maintenance!

Statistics

Security