TanStack Query is not the loudest part of TanStack’s recent public story, and that is the point. The package still shows 49,374 GitHub stars, 30 releases in 90 days, a 100 shipping score, and a 95 health score, while TanStack’s blog is busy pushing RSC and AI infrastructure around it.

The official TanStack site describes the company as an open-source application stack for the web, with headless, type-safe, composable tools for modern web applications. On that same site, Query is positioned as asynchronous state management, server-state utilities, and data fetching for TS/JS apps across React, Vue, Solid, Svelte, Angular, and Lit.

The GitHub repository frames it more narrowly and more usefully: an async state management library for fetching, caching, synchronizing, and updating server state. That is still the center of gravity.

The interesting signal is not just Query activity

ToolVitals records 6 release events in 30 days and 30 releases in 90 days for TanStack Query. That is active maintenance by any normal framework-library standard.

But the recent first-party writing is not mostly about Query release notes. It is about TanStack AI, React Server Components, TanStack Start, provider tooling, debug logging, audio generation, and test infrastructure. The React Server Components post even shows Query as part of the server-component workflow, not as a standalone product island.

That suggests a clear bet: Query remains the server-state primitive, while TanStack builds a broader application stack around routing, server rendering, AI workflows, typed provider APIs, and framework adapters.

TanStack is treating type safety as product surface area

The TanStack AI posts are useful context because they show what the team is optimizing for. The provider tools post says incompatible model and tool pairings now fail at compile time instead of being silently ignored at runtime. The debug logging post adds pipeline visibility across activities and adapters.

That does not prove Query got better because TanStack AI shipped those features. ToolVitals should not blur product lines.

It does show a consistent engineering posture across the public TanStack stack: typed APIs, provider abstraction, observable execution, and cross-framework support. Query already fits that posture. The GitHub README lists protocol-agnostic fetching, caching, refetching, pagination, mutations, dependent queries, background updates, prefetching, cancellation, and React Suspense support.

What ToolVitals cannot infer

ToolVitals can see stars, releases, release events, health, shipping, SSL, uptime, and public activity. For TanStack Query, those signals are strong: 49,374 stars, 30 releases in 90 days, 6 release events in 30 days, a 94 ToolVitals score, and 85 data confidence.

ToolVitals cannot see whether your team will enjoy the API. It cannot measure code quality inside a release. It cannot infer user satisfaction, revenue, enterprise adoption, bug severity, or whether a migration will be painful in your app.

It also cannot treat every TanStack blog post as a Query feature. Recent first-party posts prove the TanStack stack is active. They do not prove that Query itself shipped AI audio, RSC ownership models, or provider-tool typing.

Compared with nearby tools

Among related framework tools in the payload, TanStack Query has fewer 30-day release events than Next.js, which shows 24 release events and 139,347 stars. Query has more stars than Qwik, which shows 22,001 stars and 12 release events.

That comparison cuts both ways. Next.js has a larger public footprint and more recorded short-term event volume. TanStack Query looks smaller than Next.js, but more established than Qwik by stars, with a very strong 100 shipping score.

The cross-category comparisons are noisier. n8n shows 48 release events in 30 days, LangChain shows 32, and OpenClaw shows 44. Those are not direct substitutes for Query, but they set the current bar for hyperactive developer tooling. Query is active, not frantic.

Recommendation

If your frontend team is fighting server-state bugs, evaluate TanStack Query because it is still maintained like core infrastructure, not like a finished side package. The public data supports that claim: 30 releases in 90 days, 49,374 stars, and a 95 health score.

If you are choosing an application architecture, look beyond Query alone. TanStack is clearly pushing a wider stack around Start, Router, RSC, and AI. Query is the boring, proven layer in that story. Boring is good when the job is caching, refetching, synchronization, and not making production data flows weird.

Sources