The State of JavaScript 2025 survey collected responses from 13,002 developers between September and November 2025. The headline finding might surprise you: not much has changed. That's the point. After a decade of relentless churn, constant framework launches, and collective "JavaScript fatigue," the ecosystem has settled. The framework wars are effectively over.

Run annually by Devographics since 2016, the State of JS survey is one of the most reliable signals for tracking where the JavaScript ecosystem is heading. This year's edition confirms a shift that many teams have felt intuitively: the core frameworks are mature and stable, the real competition has moved to meta-frameworks and build tools, and AI is fundamentally reshaping how code gets written.

For teams building content-driven applications, this stability creates a strategic opportunity. When your frontend framework isn't going to be obsolete in 18 months, you can invest more confidently in the infrastructure around it. That's where an API-first content backend like Strapi fits in, connecting to whichever framework your team has committed to.

In brief:

• The average developer has used just 2.6 frontend frameworks in their entire career. The myth of constant framework-switching is dead.
• Astro leads meta-framework satisfaction by a 39 percentage point margin over Next.js, signaling a content-focused architecture shift.
• Vite is within 2 percentage points of overtaking webpack in adoption, while holding a massive 78-point satisfaction advantage (State of JS 2025 build tools data).
• Nearly 29% of code was AI-generated by end of 2025, a 45% year-over-year increase that's transforming development workflows.

The Ecosystem Has Reached "Peak Framework"

The State of JavaScript 2025 survey opens with a tone that captures the ecosystem's newfound calm—the era of relentless framework creation has quietly wound down.

The data backs this up in a concrete way. The average respondent has used only 2.6 frontend frameworks over their entire career. The survey editors put it directly: "The image of the burned-out web developer jumping from framework to framework on a monthly basis isn't quite accurate anymore."

The major frameworks—React, Vue, Angular, and Svelte—are all mature and stable. React, Vue, and Angular continue to serve their communities reliably. Developers aren't switching because they're bored. They're settling because the tools work.

Frontend happiness scores reinforce this. The survey notes that "while in slight decline, front-end happiness has been remarkably stable over the years," hovering around the same range consistently. Developers aren't ecstatic, but they aren't frustrated either. They've found tools that are good enough to build real things.

What This Means for Teams

The practical implication is significant for technical decision-makers. You can confidently choose a framework today knowing it will be supported and maintained for years. The question is no longer "which framework?" but "which meta-framework and content infrastructure do we build around it?"

Since frameworks are stable, the differentiator becomes what you pair them with. A headless Content Management System (CMS) like Strapi works with all the major options: React, Vue, Svelte, and Angular. Teams aren't locked into a single ecosystem when their content layer is framework-agnostic.

Frontend Frameworks: Svelte 5 Takes the DX Crown

While the overall landscape has stabilized, the satisfaction rankings tell an interesting story about where developer experience (DX) is heading.

Svelte 5 has generated significant positive attention in the survey, driven largely by its new Runes reactivity system. Runes replace Svelte 4's implicit $: reactive statements with explicit primitives: $state for reactive variables, $derived for computed values, and $effect for side effects. The result is code that's more predictable, easier to debug, and works outside component files for the first time.

Enterprise teams report real-world results. A production migration case study from Villa Plus documented a 40% reduction in update times for data-heavy dashboards, with fewer reactivity-related bugs and faster onboarding for new developers.

React remains the dominant framework by usage, but dominance comes with a cost. It's also the most common source of complaints: complexity, state management overhead, and performance concerns. Being the most popular means being the most scrutinized. There's also a structural advantage at play: React's massive codebase presence across the web means AI tools tend to generate React code most fluently, which may reinforce its dominance.

Vue maintains its steady, reliable positioning, with an approachable API and gentle learning curve that translate to fast team onboarding.

Solid deserves a mention here, too. It has maintained the highest satisfaction rating for five consecutive years (2021-2025), despite only about 10% usage. As the survey editors note: "the fact that it's had the highest satisfaction for five years running should be enough to make us pay attention to what it's doing."

The key insight: developer satisfaction doesn't necessarily correlate with market share. Choose based on your project's needs, not popularity rankings. Regardless of which frontend framework your team prefers, Strapi delivers content through RESTful and GraphQL APIs that work identically across all of them. Explore the Svelte + Strapi integration or start with the most popular pairing: React + Strapi.

The Meta-Framework Battle: Astro vs. Next.js

If the framework wars are over, the meta-framework wars are just heating up. According to the State of JS 2025 meta-frameworks data, the main battle has clearly moved to this layer.

Astro has undergone what the survey describes as a "dramatic transformation from near obscurity to serious contender." The headline number: Astro leads all meta-frameworks in developer satisfaction, holding a 39 percentage point advantage over Next.js. That's not a marginal difference.

What makes Astro compelling for content sites is its islands architecture. Pages render as static HTML by default, with zero JavaScript shipped to the browser. Interactive components opt in explicitly through hydration directives like client:load, client:idle, or client:visible. For a blog post with one interactive comment form, only that form loads JavaScript. Everything else stays as pure HTML/CSS.

Astro's "bring your own framework" approach adds flexibility that other meta-frameworks can't match. A single Astro project can use React, Vue, or Svelte components on the same page, each hydrating independently with only its required runtime.

Next.js retains clear usage leadership at an estimated 60-70% adoption, but satisfaction has declined. The increasing complexity of Server Components and the App Router has drawn real criticism from developers. The December 2025 React2Shell vulnerability (CVE-2025-55182) added urgency to these concerns, requiring patches across multiple React and Next.js versions after Palo Alto Networks confirmed active exploitation in the wild.

Nuxt holds steady as the natural companion for Vue teams. SvelteKit continues gaining traction alongside Svelte 5. TanStack Start has also emerged as a notable new entrant, achieving approximately 4% mention rate in write-in responses despite not being an official survey option due to its beta status during data collection. Its presence signals a philosophical split in the meta-framework space between integrated, full-stack solutions and composable, library-first approaches.

Developers use an average of just 1.7 meta-frameworks, so these choices tend to be sticky. Choose carefully.

Matching Meta-Frameworks to Content Types

The meta-framework decision increasingly depends on what you're building:

• Marketing sites, blogs, documentation: Astro's static-first rendering pairs naturally with API-first content delivery. The Astro + Strapi combination is gaining significant traction for exactly this use case.
• Interactive web apps: Next.js + Strapi handles app-like projects where you need SSR, ISR, and deep ecosystem support.
• Vue ecosystem projects: Nuxt + Strapi gives teams a batteries-included stack with flexible rendering strategies per route.

For a deeper comparison, Strapi has published a useful breakdown: Next.js vs Astro vs Remix: Choosing the Right Frontend Framework.

Backend Frameworks Make a Comeback

Backend frameworks returned to the State of JS survey after a three-year absence, a recognition of JavaScript's growing role on the server.

Express remains the usage leader, but the survey characterizes it as "reflecting its age in satisfaction metrics." High adoption, moderate satisfaction. Newer frameworks tell a different story.

Hono, Nitro, and ElysiaJS are gaining satisfaction and interest, especially among developers seeking modern, TypeScript-friendly options. Hono stands out for edge and serverless use cases: it's built entirely on Web Standard APIs, ships at roughly 14KB minified, and runs on Cloudflare Workers, Deno, Bun, and Node.js without code changes. Its multi-router architecture (five specialized routers optimized for different patterns) targets the resource-constrained environments where Express simply doesn't fit.

ElysiaJS is closely associated with Bun as a runtime. That alignment looks increasingly validated after Anthropic's acquisition of Bun in December 2025, signaling serious investment in the JavaScript runtime for AI infrastructure.

NestJS continues growing in enterprise environments where structured, opinionated architecture is an advantage rather than a constraint.

Beyond frameworks, the emergence of typed APIs via tools like tRPC and oRPC is reshaping how frontend and backend communicate, enabling end-to-end type safety that reduces integration bugs and improves developer confidence.

For teams combining structured content management with custom API logic, a Node.js-based CMS like Strapi can complement these backend frameworks by providing a content API layer that works alongside them. You can learn more about Strapi's features for backend content management.

Build Tools: Vite's Inevitable Takeover

The build tools data from the State of JS 2025 tells a clear story. Webpack maintains a narrow usage lead at 86.4% versus Vite's 84.4%, a gap of just 2 percentage points. But satisfaction tells the real story: Vite achieves 56% positive sentiment with only 1% negative (net +55), while webpack shows 14% positive against 37% negative (net -23). That's a 78-point satisfaction differential.

Configuration complexity was the #1 pain point cited by build tool users, with 289 respondents flagging it explicitly. If a tool doesn't come with a zero-config CLI in 2025, adoption drops sharply.

The next chapter is already being written in Rust. The VoidZero initiative unifies Rolldown (a Rust-based bundler) and Oxc (Rust-based compiler and linter) into a cohesive toolchain. Rolldown achieved an 80% interest ratio in the survey and is on track to become Vite's internal bundler by 2026. This isn't a Vite replacement; it's Vite's evolution. The developer-facing API stays the same while the internals get dramatically faster.

Pairing Strapi with Vite-powered frontends (via Astro, SvelteKit, or Nuxt) creates fast, developer-friendly workflows from content creation to deployment.

AI's Impact: 29% of Code Is Now AI-Generated

Nearly 29% of code produced by survey respondents was AI-generated by the end of 2025.

That's up from 20% the previous year, a 45% relative increase, according to the State of JS 2025 usage data. The survey conclusion characterizes this as "just under 30%." The survey also found that roughly 30% of developers reported writing less code by hand than the previous year, underscoring the practical impact of these tools on daily workflows.

The AI tools data reveals which tools developers actually use:

1. ChatGPT: 6,519 respondents
2. GitHub Copilot: 5,584 respondents
3. Claude: 4,845 respondents
4. Google Gemini: 3,074 respondents
5. Cursor: 2,829 respondents

On the editor front, VS Code remains dominant at 9,186 respondents, but AI-native editors are carving out real territory. Cursor (2,797 respondents) and Zed (1,387 respondents) combine for over 4,100 users choosing editors specifically designed with AI-first workflows.

The shift from reactive code completion to autonomous agents is accelerating. GitHub Copilot now includes an autonomous agent mode for multi-step task execution. Claude Code operates with full system access to plan, code, test, and iterate. These aren't autocomplete tools anymore. The broader trend toward tool-use protocols like the Model Context Protocol (MCP) suggests AI coding assistants will increasingly interact directly with development infrastructure, from databases to deployment pipelines.

What This Means for Developer Skills

The skills that matter in 2026 are shifting from syntax knowledge to system design, code review, and understanding how components fit together. Framework choice matters less when AI handles the syntax fluently. What matters is knowing whether the generated code is secure, accessible, and architecturally sound.

As AI tools generate more frontend code, having a structured, API-first content backend becomes even more critical. Strapi provides the structured data layer that AI-generated frontends can reliably consume. Its REST and GraphQL APIs provide consistent, well-documented endpoints that AI tools can integrate with regardless of which framework they scaffold. You can explore available Strapi plugins on the marketplace to extend this functionality further.

What This Means for Your Stack in 2026

Here's how to translate these findings into practical decisions for different team profiles:

• Content-heavy sites (blogs, marketing, docs): Astro + Strapi is the emerging standard. Static-first rendering plus headless CMS equals performance and editorial flexibility without shipping unnecessary JavaScript.
• Interactive web apps: React/Next.js remains the safest bet for hiring depth and ecosystem maturity. Pair with Strapi for content management to keep your application code focused on business logic. For teams exploring alternatives, the Remix + Strapi integration offers a server-first architecture worth evaluating.
• Performance-critical projects: Svelte 5 + SvelteKit offers strong DX and runtime performance, with community benchmarks showing meaningful advantages in bundle size and update speed. Strapi integrates via its API layer without adding framework-specific overhead.
• Vue ecosystem teams: Nuxt + Strapi gives you flexible rendering strategies (SSR, SSG, ISR per route) with excellent developer experience.
• Legacy or static-first projects: Gatsby + Strapi remains a solid option for teams invested in Gatsby's GraphQL-based static generation pipeline.
• All teams: Adopt Vite if you haven't already. Invest in AI tooling literacy. Default to TypeScript. And choose tools that are framework-agnostic for your content layer, so your infrastructure decisions outlast your framework preferences.

The Post-Framework Era: Building for Stability, Flexibility, and AI-Driven Development

The State of JavaScript 2025 paints a picture of an ecosystem that has grown up. Less fear of missing out, more pragmatism. Developers are building with tools they trust, not chasing the newest thing.

The real question for 2026 isn't "which framework?" It's "how do I build a content infrastructure that adapts regardless of what changes?" As AI reshapes how code gets written and frameworks continue to stabilize, the tools that survive will be the ones that stay flexible, API-first, and framework-agnostic.

Explore how Strapi connects with every major framework covered in this article at strapi.io/integrations.

All data in this article is sourced from the State of JavaScript 2025 survey, conducted by Devographics between September and November 2025 with 13,002 respondents.

Get Started in Minutes

Run npx create-strapi-app@latest in your terminal and follow our Quick Start Guide to build your first Strapi project.

• View Quick Start Guide