Strava imposes $11.99 API fee and restricts data access ahead of IPO
CEO Michael Martin cites unchecked data harvesting by AI labs and server overload as primary drivers for the new measures, which include a flat monthly developer fee and mandatory authentication for public profiles.
Fitness and social running platform Strava has announced a significant overhaul of its data access policies, introducing a flat monthly fee of $11.99 for developer API access and restricting website visibility to authenticated users. The changes, which include the retirement of certain API endpoints and the adoption of the Model Context Protocol (MCP), are designed to curb unauthorised data scraping by artificial intelligence companies ahead of the firm’s planned initial public offering.
The move replaces Strava’s previous free, tiered access program with a uniform subscription model, although the company noted that pricing may vary by geography. This shift comes as the platform’s developer community has expanded from 185,000 members last year to 241,000 this year. To assist with the transition, Strava has provided a 90-day grace period for developers to adapt to the new rules.
Strava is also tightening security around its website, requiring users to log in to view data that was previously public, such as user profiles and fitness club listings. Additionally, the company plans to retire specific API endpoints, including those providing access to club details, to further protect user data. The introduction of the Model Context Protocol will allow Strava to maintain greater control over how AI assistants and applications access external data in a structured manner.
Chief Executive Officer Michael Martin cited unchecked AI scraping and server overload as the primary drivers for these changes. Martin described the aggressive harvesting of public websites by AI companies as potentially the "death knell of the public internet," noting that the endless demand for training data has degraded site performance and impaired system stability. He also highlighted server strain caused by poorly structured applications generated by "vibe coding" tools.
Strava has refused data licensing deals from major AI labs, with Martin specifically singling out Perplexity for routing its scraping through aggregator services to obscure its origin despite being turned away. The strategy mirrors similar actions by other tech firms, such as Meta’s ban on third-party chatbots and Reddit’s 2024 API crackdown, though Strava’s flat fee structure aims to keep the developer ecosystem intact unlike Reddit’s call-based pricing model.
