Launching Brellium – One API for All AI Models

With new AI models coming out every day, I’m constantly playing with new models to see which best fits my projects. For a while, Davinci was my default. But now, I toggle between Claude, 3.5 Turbo, Llama, 4, and whatever’s out next to figure out the best model to use for each project.

Each API requires different configuration, prompting, and setup. That’s OK once, but with new models coming out almost daily, I’m spending half my time tweaking and translating my prompts to each new model.

So, we built Brellium (https://www.brellium.com) - it’s a single endpoint that aggregates the most popular models. Meaning, you can now test, ship, and iterate on new models in a fraction of the time.

You simply make a cURL request to our endpoint and get back responses from each popular endpoint, data points like cost and latency, plus a recommended query based on what similar queriers have chosen.

If you have any specific recommendations, just comment below. Would love to hear what you guys think!