CRE-Router

Implementation of the paper Cluster, Route, Escalate: Cascaded Framework for Cost-Aware LLM Serving.

Production LLM serving trades accuracy against cost. CRE-Router routes each query to the most cost-effective model in a pool, then escalates low-quality outputs to a stronger model:

• Stage 1 (clustering-based routing). Queries are embedded and clustered offline; each cluster is routed to the model that minimizes a cost-adjusted error score (Error + lambda * Cost), with lambda tuned once to satisfy a latency (TPOT or E2EL) budget.
• Stage 2 (quality-estimation cascade). A lightweight ModernBERT classifier inspects each efficient-model output and escalates low-quality answers to the most accurate model the routing uses. cre fit plans which clusters to gate.

Both stages train only on task-correctness labels obtainable from standard benchmark evaluation; no extra annotation is required.

Install

pip install "cre-router[full]"

full installs the whole pipeline on one machine. Narrower installs are available through the serve, qe, eval and data extras; see the installation guide.

Usage

The workflow is driven by the cre CLI, one stage per step:

cre cluster → cre evaluate → cre fit → cre qe-data → cre qe-train → cre serve

Offline, from saved captures: cre stats → cre fit → cre compose.

Run cre <stage> --help for options. Runnable end-to-end quickstarts (a no-GPU routing-table demo and a full serving walkthrough) are in the README.

Links

• Source and documentation: https://github.com/ymoslem/CRE-Router
• Paper: https://arxiv.org/abs/2606.27457
• Reproducing the paper: https://github.com/ymoslem/CRE-Router/blob/main/REPRODUCE.md

Apache-2.0 licensed.