| 开发者 | cubixsol |
|---|---|
| 更新时间 | 2026年8月1日 00:03 |
| PHP版本: | 7.4 及以上 |
| WordPress版本: | 7.0 |
| 版权: | GPLv2 or later |
| 版权网址: | 版权信息 |
{title}, {prompt}, and {style}. Every generated asset lands in your database completely optimized for Google Image Search, completely eliminating unindexed filenames like IMG_4382.png.
💰 Granular Budget Control
aiisp_register_providers, aiisp_register_stock_sources, and aiisp_style_presets.
No. The core plugin architecture is entirely free. The integration with Pollinations.ai remains keyless and unmetered. For premium engines, you utilize your own directly managed developer accounts—many of which feature generous free usage tiers (such as Google Gemini and Leonardo.Ai). You pay only for what you generate, directly to the provider, without third-party markups.
Absolutely not. The plugin is built following strict performance guidelines. Script initialization and execution occur exclusively within wp-admin workflows under specific context checks. No assets are queued or rendered on the public-facing side of your site.
Your API keys are securely stored in your isolated site database via the standard WordPress Options API. They are never transmitted back to visual form fields after validation, nor are they ever shared with external tracking networks. Communication occurs solely between your host server and the specific engine endpoint.
The test interface initiates a minimal query to the platform's authentication server. Where standard key validation routines exist, it returns a 0-cost structural check. For engines that validate exclusively during generation tasks (such as Ideogram), the system executes a microscopic, fraction-of-a-cent generation utilizing their most economical model to guarantee clear end-to-end functionality.
If your active engine encounters a structural error or experiences timeout intervals, the plugin automatically falls back down your defined priority stack to process the request seamlessly. Every failover event is captured and indexed on the administration dashboard for architectural review.
Yes. The plugin is explicitly built to be extended. Engines, asset sources, and style lists are built on standard registries. By binding custom classes to aiisp_register_providers, teams can easily add proprietary models or customized internal API connections without modifying the core files.
gpt-image-2 integration following provider API adjustments (May 12, 2026). Updated global dimension validation sets.response_format within OpenAI pipelines caused by structural mutations on the vendor endpoint.