A research paper introduces ensemble routing, a method that lets Web-based LLM services select and combine multiple models per prompt rather than committing to one model or a fixed ensemble, so that response quality and inference cost can be balanced across heterogeneous requests. The approach is presented as adapting ensemble composition and cost to each individual request, addressing the limitation that conventional routing picks a single model per prompt while fixed ensembles cannot adjust to individual requests.
Large language models (LLMs) increasingly support Web-based services that must balance response quality and inference cost across heterogeneous requests. Conventional routing selects one model per prompt, while fixed ensembles cannot adapt their composition and cost to individual requests. We introd
EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.