Report card

Do AI agents pick LaunchDarkly?

How often agents choose LaunchDarkly when a developer needs feature flags — measured across Claude, GPT, Gemini.

LaunchDarkly

#1 of 5 feature flags
25%Pick Rate
95% CI 20%31%

When developers ask an AI agent for feature flags, LaunchDarkly is picked 25% of the time — ranking #1 of 5 measured across Claude, GPT, Gemini.

Picked 25% Named, not picked 44% Never surfaced 31%

Top pick in the category. 🟢

Awareness

1 docs in FineWeb (10BT sample) · 1.5K URLs in Common Crawl

Present in the raw crawl and the filtered corpus models train on — solid corpus footprint.

AdoptionDid the pick convert to installs? Public Adoption = npm downloads of your packages. The proof half of the funnel.

17.1M installs / month ▼ 4% MoM

Installs are roughly flat month over month.

Get the full report

The per-model and per-surface breakdown for LaunchDarkly — where it wins, where it loses, and to whom — plus an alert when the Pick Rate moves.

About LaunchDarkly

A feature management platform for controlling feature rollouts without redeploying code.

LaunchDarkly lets development and product teams enable or disable features for specific users or percentages of traffic at runtime, through a central dashboard. It supports gradual rollouts, targeted access by user attributes, and instant kill switches for features causing problems in production. It is aimed at teams running web services and multi-user applications who want to separate deployment from feature release.

Where to start

Start by creating a LaunchDarkly account and following the SDK reference guide for your language to initialize the client with your SDK key and evaluate flags in your application code.

Install

Links and summary verified from public sources.

The rest of the Feature Flags ranking

← Full Feature Flags leaderboard · How this is measured