Stripe acquires AI start-up OpenRouter for $7.5B
What's the story
Stripe, a leading player in the payments industry, has announced its decision to acquire artificial intelligence (AI) start-up OpenRouter. The acquisition is part of Stripe's strategy to enter the AI model market. Although financial details of the deal have not been disclosed, reports suggest that it could be worth around $7.5 billion. The acquisition comes just three months after OpenRouter raised $113 million at a valuation of about $1.3 billion.
Developer appeal
OpenRouter's AI models are free and non-proprietary
OpenRouter has gained a lot of traction among developers looking to use AI models, especially those that are non-proprietary and free.
These open-weight AI models mostly come from Chinese labs like DeepSeek and Z.ai.
They have become popular for being more cost-effective than proprietary AI models from US companies such as OpenAI and Anthropic.
Cost management
Stripe's blog post on AI model usage
In a blog post about the acquisition, Stripe revealed that it has been working with companies to "optimize their token costs and route tokens efficiently." This refers to a metric used to measure AI model usage.
The company also noted that managing AI costs relative to performance is challenging due to the rapid "pace at which models are released and repriced."
Market position
Stripe's growth and recent acquisition spree
Stripe is one of the world's most valuable start-ups, with a valuation of nearly $160 billion earlier this year. The company's online payment technology has become ubiquitous in many markets.
Last year, it expanded its crypto exposure by acquiring stablecoin platform Bridge for $1.1 billion.
Now, with the acquisition of OpenRouter, Stripe is further solidifying its position in the AI market, which is much bigger and growing at a faster pace.
Future goals
OpenRouter's vision for a healthy AI ecosystem
OpenRouter has said that its acquisition by Stripe will help realize its vision of "a healthy AI ecosystem where many models thrive."
The company wants this diversity to be a strength, allowing labs or inference providers with breakthroughs to reach millions of developers.
It also hopes "no single model becomes the default by inertia."