Last update: September 12th, 2025
<aside> 💡
In short: Our tech stack evolves constantly. We’ll keep this page updated, but exact model versions or database setups may change as we continue experimenting.
</aside>
At Shipper, we don’t lock ourselves into just one model. Instead, we’re always experimenting to see what works best for our users.
We regularly test and combine models from:
Sometimes the newest models give us the best results, and sometimes the older, more proven ones are more reliable for certain tasks. We like to keep a balance between cutting-edge and battle-tested.
The same goes for our infrastructure. On the database side, we safely and securely try out new technologies while continuing to rely on stable, trusted systems. That way, your projects get both performance and reliability.
For those interested in the underlying tech, here’s a closer look at how we approach models and databases at Shipper:
We run experiments across a mix of cutting-edge and stable large language models. At the moment, our stack may include: