Laravel Ai Laravel Package

Product Decisions This Supports

• Feature decisions around image-generation safety and compliance: v0.12.3 lets product/engineering pass a caller-supplied moderation value through PendingImageGeneration::moderation() into the OpenAI Responses API image_generation tool. PM can tune moderation posture per feature, audience, or market without forking the package.
• Build vs. buy: the package provides a Laravel-native provider abstraction, fluent image-generation API, and test/fake gateway, reducing bespoke integration work. This release closes a parity gap between the Responses API and classic Images API path for moderation control.
• Roadmap: enables use cases such as consumer-facing image generation with stricter controls, internal/enterprise tools with relaxed controls, regional compliance variants, and UGC/marketing asset generation where brand-safety settings must be explicit.
• Vendor strategy: OpenAI Responses API is the only path that honors the new parameter; other gateways accept it for interface compatibility but ignore it. PM should treat moderation control as OpenAI-specific, not a cross-provider guarantee.
• Risk/ops: the package imposes no default on any path; when omitted, OpenAI applies its own default. The classic Images API path still forces moderation: 'low' for gpt-image*, so product behavior differs by API path.

When to Consider This Package

• Consider it when building a Laravel app that needs image generation with OpenAI's Responses API and wants configurable moderation without custom HTTP plumbing.
• Good fit if you value a fluent, testable integration (moderation() mirrors quality()), backward-compatible additions, and an abstraction that can route to multiple providers.
• Also consider if you already use this package for other AI features and want to extend image generation with explicit safety controls.
• Look elsewhere or plan custom work if you need consistent moderation semantics across all providers, because Gemini, OpenRouter, Azure OpenAI, Xai, Bedrock, Anthropic, and the fake gateway ignore the parameter.
• Look elsewhere if you require caller-controlled moderation on the classic Images API path for gpt-image* models; that path still forces moderation: 'low'.
• Be cautious with a v0.x dependency if you need long-term API stability; pin versions and review changelogs. If you are not on Laravel/PHP, this package is not relevant.

How to Pitch It (Stakeholders)

• Executives: v0.12.3 is a small, backward-compatible feature that improves safety and compliance control for AI-generated images while avoiding custom integration cost. It supports differentiated product experiences by audience or market and reduces vendor lock-in through a Laravel-native abstraction.
• Engineering: the release adds an optional ?string $moderation to ImageGeneration, threads it through PendingImageGeneration, ImageProvider, ImageGateway, and GeneratesImagesViaResponses, and includes it only when provided—matching the existing quality pattern. No breaking changes.
• Product/risk: moderation control is currently OpenAI Responses API-specific; do not promise cross-provider moderation parity. The classic Images API path remains unchanged and forces moderation: 'low' for gpt-image*. Treat this as an incremental capability, not a complete moderation strategy.