Meta just shipped a vibe-coded app to the App Store. Vibe coding isn't a hobbyist story anymore.

Meta quietly launched Pocket, a consumer app built on its Gizmo acquisition that lets anyone generate small playable games — 'gizmos' — from a text prompt, no code involved. When a company Meta's size ships prompt-generated software to production, the debate about whether AI-built apps are 'real' is effectively over.

Meta quietly launched Pocket on the App Store and Google Play on 29 June, without an official announcement. It’s a creative platform for making and sharing “gizmos” — small interactive games and apps generated from a text prompt — built on the team Meta acquired from vibe-coding platform Gizmo earlier this year. Users scroll a feed of gizmos other people made and can remix or build their own with no coding involved.

Why a quiet launch from Meta is the real signal

Vibe coding has spent the last two years as a story about solo founders and weekend prototypes. Pocket is different: it’s a production consumer app, shipped by one of the largest software companies in the world, built entirely on prompt-generated content as the product itself rather than a scaffold a developer then rewrites. Meta didn’t build Pocket to prove a point about AI — they built it because the economics and the output quality finally justified shipping it to hundreds of millions of users. That’s a different signal than another startup’s funding round.

The catch that still applies

Nothing about Pocket contradicts what we’ve said before about AI-generated software: it works well for small, contained, disposable experiences where a bug means a bad game, not a bad transaction. Games and gizmos tolerate imperfection in a way that a checkout flow, a patient record, or an appointment booking system does not. Meta shipping this doesn’t mean the same prompt-to-production pipeline is safe for a regulated, revenue-critical, or data-sensitive product — it means the bar for what AI-generated software is trusted with just moved, and founders evaluating “should we just vibe-code our MVP” now have a genuine Big Tech data point to weigh, not just anecdotes.

So what

If you’re deciding whether a prompt-built prototype is enough or whether you need an engineered product, the honest answer depends on what happens when it’s wrong — not on whether the technology can generate it. Our custom software development page covers where we draw that line in practice, and get in touch if you want a straight read on your specific case.

← All research

More research.

GitHub's Copilot metrics now count OpenCode and Claude Code traffic: who really uses which AI coding tool

On 23 October 2026 GitHub added a per-tool breakdown to the Copilot usage metrics API, so tools such as OpenCode and Claude Code are counted even without editor telemetry. It shows how mixed the AI coding tool stack inside software teams has become.

Read ›

GitHub Copilot retires Claude Opus 5 and GPT-5.6 Terra on 16 November: the model-churn bill for buyers

GitHub has announced that Claude Opus 4.8, Claude Opus 5 and GPT-5.6 Terra leave every Copilot experience on 16 November 2026. For anyone commissioning AI-assisted development, model retirement is now a monthly event to plan around.

Read ›

GitHub Copilot now benchmarks how 'agent-first' your developers are against your industry

On 22 October 2026 GitHub added opt-in industry comparisons to its Copilot impact dashboard, ranking firms by share of agent-first developers and by the merge rate of agentic pull requests. Here is how to read the numbers when you are buying software.

Read ›

Apple's new bait-and-switch rule 2.5.19: what it means for AI-built and fast-shipped apps

On 16 October 2026 Apple updated its App Review Guidelines: apps that change after approval can be removed from the App Store and the Developer Program, and AI-generated intimate imagery is now explicitly covered. Here is what founders commissioning an iOS app should check.

Read ›

GitHub Copilot agentic autofix is generally available: AI now fixes security alerts, and bills for it

GitHub made agentic autofix for code scanning alerts generally available on 15 October 2026. Copilot can now fix up to 25 security alerts in one pull request, using AI credits your supplier or your team pays for.

Read ›

GitHub AI Scan for pull requests goes GA: AI security checks for PHP, Terraform and Dockerfiles

GitHub's AI Scan for pull requests became generally available on 15 October 2026, checking languages the standard scanner misses. Findings are advisory and cannot block a merge, which matters when you buy software.

Read ›

Start a project

Building with AI? Let's talk.

Tell us what you are building. We'll give you an honest view of the right approach.

Start a projectSee our work