The 70-Point Divide
WebPulse scores every major web framework on AI readiness — the ability to serve structured data to AI agents, expose API-first interfaces, operate headlessly, and return machine-readable responses. FastAPI scores 95. Joomla scores 25. The distance between the most AI-ready framework and the least AI-ready framework is 70 points on a 100-point scale. This is not a nuanced continuum. It is a binary: frameworks that were architected for machine consumption, and frameworks that were not.
Three Tiers, Two Gaps
The data reveals three distinct tiers. The seven recommended frameworks — FastAPI (95), Astro (92), Next.js (88), Hugo (88), SvelteKit (85), Nuxt.js (82), and HTMX (78) — average 86.9 in AI readiness. The five acceptable frameworks — Spring (80), Django (75), Laravel (74), Rails (72), and Flask (70) — cluster between 70 and 80. The three caution-rated frameworks — WordPress (35), Magento (30), and Joomla (25) — average 30.0.
The gap between the recommended average (86.9) and the caution average (30.0) is 56.9 points. But the more telling number is the void between the acceptable tier and the caution tier. Flask, the lowest-scoring acceptable framework, scores 70. WordPress, the highest-scoring caution framework, scores 35. That is a 35-point chasm where no framework exists. Frameworks either clear 70 or they fall below 35. There is no middle ground.
Architecture, Not Age — But Age Is a Near-Perfect Proxy
AI readiness measures structural capability: does the framework natively output structured data? Is the architecture API-first? Can it operate headlessly — serving content without rendering HTML? Does it return machine-readable responses by default? These are architectural decisions baked into a framework's core design, not features that can be bolted on through plugins. The correlation with framework age is not causal, but it is nearly perfect. Frameworks built after 2018 — FastAPI, Astro, SvelteKit, HTMX — were designed in an era where APIs were the primary interface and server-rendered HTML was optional. Frameworks built before 2010 — WordPress, Joomla, Drupal — were designed to generate HTML pages for human browsers. The architectural assumptions of the era determined the AI readiness ceiling.
7.4 Million Sites at 35
WordPress scores 35.0 in AI readiness and powers 7,427,780 detected sites. That is 7.4 million web properties that are structurally unprepared for AI agent traffic — the fastest-growing category of web consumption in 2026. These sites return server-rendered HTML with embedded navigation, sidebars, footers, and advertising scripts. An AI agent parsing a WordPress page must extract meaning from a document designed for human visual scanning. An AI agent querying a FastAPI endpoint receives exactly the structured data it requested, in the format it specified.
The business exposure is not theoretical. AI agents already account for a significant share of web traffic. Organizations whose sites score below 35 in AI readiness are invisible to this traffic. Their content is not indexed by AI systems. Their products are not surfaced in AI-generated recommendations. Their services are not discoverable through agentic browsing. A 35 AI readiness score is not a low grade — it is structural exclusion from the AI-mediated web.
The Acceptable Tier Is the Decision Point
The five acceptable-tier frameworks — Spring (80), Django (75), Laravel (74), Rails (72), Flask (70) — occupy the strategic middle ground. They were not designed API-first, but they have mature API capabilities. They generate HTML by default, but they support headless operation. Their AI readiness scores cluster between 70 and 80, placing them closer to the recommended tier than to the caution tier. Organizations running these frameworks face a modernization path, not a rebuild. The distance from Django at 75 to Next.js at 88 is 13 points. The distance from WordPress at 35 to Django at 75 is 40 points.
For executives evaluating their web infrastructure, the question is which side of the cliff their framework sits on. Above 70, the path forward is incremental improvement. Below 35, the path forward is architectural replacement. The 35-point void between these tiers is the clearest signal in the data: there is no gradual migration from the caution tier to the acceptable tier. The gap is too wide. The architecture is too different. The cliff is real.
