A privacy-focused, cost-aware LLM router providing unified text API access through one key.

We built Basktre to reduce the complexity of integrating multiple LLM providers. One API key provides unified text access to the live model catalog, while optional Auto Mode routes each request to a cost-effective model suited to the task. Developers can also pin a specific model whenever consistency matters. Basktre does not store prompt or response bodies. Pricing is pay-as-you-go at displayed per-token rates, with a 4% wallet top-up fee and no subscription or seat costs. The beta currently lists 10 models across 3 providers, and new accounts receive $1 in launch credits with no credit card required. We would especially value feedback from developers currently managing multiple provider keys, SDKs, and model-selection logic.

finally a router that does not feel like it is guessing in the dark. the latency numbers were noticeably better than my old setup, and the privacy stance actually shows up in the docs instead of just being a marketing line. solid.

@obanselim68454 Appreciate it, Selim! Stoked that you noticed the latency boost and appreciated the actual documentation.

Definitely let me know what you think needs changing or improving as you test it out more!

@obanselim68454 Appreciate it, Selim! Stoked that you noticed the latency boost and appreciated the actual documentation.

Definitely let me know what you think needs changing or improving as you test it out more!

Basktre

Privacy-focused, cost-aware routing for LLM APIs

Privacy-focused, cost-aware routing for LLM APIs