How Claris Uses AI in Custom App Development

Claris builds AI into its platform in two main ways: AI script steps and AI functions inside Claris FileMaker, plus LLM integration that lets custom apps call large language models directly. The company also markets an "agentic development" approach and a semantic image search capability. These features are aimed at teams already building on FileMaker who want to add AI-driven behavior without leaving the platform. If you are evaluating Claris for AI work, the practical question is which of these capabilities fits your app, and that depends on whether you need text generation, semantic search, or development assistance.

AI script steps and AI functions in FileMaker

Claris documents two distinct building blocks, both hosted under its Pro Help documentation:

• AI script steps — automation actions you add to a FileMaker script, so an app can send a prompt or request to a model as part of a workflow.
• AI functions — calculation-level functions you call inside formulas, so AI output can feed directly into fields, layouts, or logic.

The split matters because it maps to how FileMaker developers already work. Script steps handle sequenced, event-driven behavior (a button press, a record creation, a scheduled task). Functions handle inline computation. If your goal is "when a user submits a form, summarize the notes field," that is a script step. If your goal is "show a sentiment score next to each record," that is closer to a function.

Claris frames the broader capability as "Supercharge custom apps with AI tools and LLM integration," which confirms that LLM connectivity is part of the same feature set rather than a separate product.

LLM integration and semantic image search

Beyond the two primitives, the Claris site lists:

• Work with LLMs — connecting FileMaker apps to large language models.
• Semantic image search — searching images by meaning rather than filename or tag.
• AI Support — a support category for AI-related questions.

Semantic image search is the more specialized item. Traditional FileMaker image handling relies on stored metadata or container fields; semantic search implies the app can match an image to a natural-language query. That is useful for media libraries, inspection photos, or product catalogs where manual tagging does not scale.

A quote featured on the Claris homepage captures the company's stated rationale: "The AI development ecosystem is moving fast. The fastest and most powerful way to bring that capability to our community is by meeting AI where it already lives." In practice, that means Claris is integrating AI into FileMaker rather than asking customers to adopt a separate AI product.

Agentic development

Claris labels a newer capability "Agentic development — Build more with AI on your side." The site positions this alongside release highlights rather than as a standalone tool. The intent is that AI assists the development process itself, not only the apps you ship. Treat this as a direction Claris is investing in; the homepage does not spell out the mechanics, so if agentic development is central to your decision, check the release notes and documentation linked from the Claris resources hub before committing.

How to decide what to use

Two practical caveats. First, the Claris site separates Pricing and Start free as distinct entry points, and lists upgrade as a related signal — so confirm current plan terms on the pricing page rather than assuming any AI capability is included at every tier. Second, the AI features are documented in the Pro Help section, which is the authoritative place to check exact syntax and availability for your FileMaker version.

If your app already lives in FileMaker, the lowest-friction starting point is a single AI script step in an existing workflow — for example, routing incoming support notes through a summarization step — and expanding from there once you have confirmed model access and cost behavior.