Strava closes its API to AI apps: $11.99 a month for developers

Strava, the fitness platform used by millions of athletes, is introducing a paid subscription for developers who want to work with its API. The price is $11.99 a month. The reason: an explosive rise in scrapers and AI tools that strain its servers and routinely violate the terms of use.

Why the API was closed

Strava has restricted free API access and put a paywall in front of developers. It is a response to a wave of unwanted apps built in minutes on zero-code platforms, without serious programming. The company complains that such tools make it easy to build apps that flood the API with requests and collect user data without checking compliance with the rules.

The company also points to "API intermediaries": middlemen who broke the terms of the agreement and resold access to the data. Scrapers degraded the platform for all users by slowing down standard operations.

The scale of the problem

Strava disclosed figures showing the escalation:

• Developer applications grew 448% over the past year
• Intermediaries (API intermediaries) regularly violated the license terms
• Scrapers and bots slowed the platform for legitimate users
• Zero-code builders let people create apps without deep programming knowledge

It is a typical pattern: once a platform becomes popular, scrapers and data speculators start hunting it. This is especially true when the data is valuable, and Strava's fitness data, with its routes, times and personal records, is very valuable for training AI models.

Who this affects

Startups and developers use fitness-tracker APIs to assemble datasets for AI. Strava is one of the largest sources of physical-activity and geolocation data, which attracts both legitimate apps (weather apps, social networks for runners) and scrapers.

"We are feeling this in full: applications grew 448% year over year,

intermediaries violated the terms, and scrapers spoiled the experience for all users," the company explains in an update on its website.

The paid model will stop legitimate developers who cannot spare $12 a month, but scrapers will find workarounds: parsing the web version, using browser bots, bypassing captchas.

What it means

The wall around data is getting higher. After the AI industry devoured most of the open internet, platforms are starting to protect their datasets with payments and access restrictions. This will not fully stop scrapers, which will find workarounds. But it will create friction for legitimate developers and startups that have no budget to negotiate with API providers.

For large companies like Apple or Google, this is no obstacle: they pay and sign contracts. For solo developers and early-stage startups, it is one more barrier to the data they need to train models.