# RouteNLP: Closed-Loop LLM Routing with Conformal Cascading and Distillation Co-Optimization

arXiv: https://arxiv.org/abs/2604.23577 · HTML: https://mlsanity.com/paper/2604.23577/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 27 of its 37 references that are in the map.
- First posted on arXiv 26 Apr 2026.

## Builds on (most cited first, 27 of 27)

- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://mlsanity.com/paper/1910.01108/) (2019)
- [Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer](https://mlsanity.com/paper/1701.06538/) (2017)
- [Self-Refine: Iterative Refinement with Self-Feedback](https://mlsanity.com/paper/2303.17651/) (2023)
- [GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers](https://mlsanity.com/paper/2210.17323/) (2022)
- [Fast Inference from Transformers via Speculative Decoding](https://mlsanity.com/paper/2211.17192/) (2022)
- [AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration](https://mlsanity.com/paper/2306.00978/) (2023)
- [BloombergGPT: A Large Language Model for Finance](https://mlsanity.com/paper/2303.17564/) (2023)
- [FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performance](https://mlsanity.com/paper/2305.05176/) (2023)
- [Medusa: Simple LLM Inference Acceleration Framework with Multiple Decoding Heads](https://mlsanity.com/paper/2401.10774/) (2024)
- [Sequence-Level Knowledge Distillation](https://mlsanity.com/paper/1606.07947/) (2016)
- [RouteLLM: Learning to Route LLMs with Preference Data](https://mlsanity.com/paper/2406.18665/) (2024)
- [Efficient Intent Detection with Dual Sentence Encoders](https://mlsanity.com/paper/2003.04807/) (2020)
- [Hybrid LLM: Cost-Efficient and Quality-Aware Query Routing](https://mlsanity.com/paper/2404.14618/) (2024)
- [Distribution-Free, Risk-Controlling Prediction Sets](https://mlsanity.com/paper/2101.02703/) (2021)
- [CUAD: An Expert-Annotated NLP Dataset for Legal Contract Review](https://mlsanity.com/paper/2103.06268/) (2021)
- [Large Language Model Routing with Benchmark Datasets](https://mlsanity.com/paper/2309.15789/) (2023)
- [S-LoRA: Serving Thousands of Concurrent LoRA Adapters](https://mlsanity.com/paper/2311.03285/) (2023)
- [Large Language Model Cascades with Mixture of Thoughts Representations for Cost-efficient Reasoning](https://mlsanity.com/paper/2310.03094/) (2023)
- [Language Model Cascades: Token-level uncertainty and beyond](https://mlsanity.com/paper/2404.10136/) (2024)
- [Fly-Swat or Cannon? Cost-Effective Language Model Choice via Meta-Modeling](https://mlsanity.com/paper/2308.06077/) (2023)
- [A Unified Approach to Routing and Cascading for LLMs](https://mlsanity.com/paper/2410.10347/) (2024)
- [Model Cascading: Towards Jointly Improving Efficiency and Accuracy of NLP Systems](https://mlsanity.com/paper/2210.05528/) (2022)
- [MetaLLM: A High-performant and Cost-efficient Dynamic Framework for Wrapping LLMs](https://mlsanity.com/paper/2407.10834/) (2024)
- [FrugalML: How to Use ML Prediction APIs More Accurately and Cheaply](https://mlsanity.com/paper/2006.07512/) (2020)
- [Automatically Adaptive Conformal Risk Control](https://mlsanity.com/paper/2406.17819/) (2024)

## Built on by (most cited first, 2 of 2)

- [HyDRA: Hybrid Dynamic Routing Architecture for Heterogeneous LLM Pools](https://mlsanity.com/paper/2605.17106/) (2026)
- [Uncertainty Quantification for LLM Agents: A Taxonomy, an Evaluation Protocol, and an Empirical Study](https://mlsanity.com/paper/2609.07395/) (2026)

## Abstract (arXiv)

Serving diverse NLP workloads with large language models is costly: at one enterprise partner, inference costs exceeded 200K/month despite over 70% of queries being routine tasks well within the capability of smaller models. We present RouteNLP, a closed-loop framework that routes queries across a tiered model portfolio to minimize cost while satisfying per-task quality constraints. The framework integrates three components: a difficulty-aware router with shared task-conditioned representations trained on preference data and quality signals; confidence-calibrated cascading that uses conformal prediction for distribution-free threshold initialization; and a distillation-routing co-optimization loop that clusters escalation failures, applies targeted knowledge distillation to cheaper models, and automatically retrains the router, yielding over twice the cost improvement of untargeted distillation. In an 8-week pilot deployment processing ~5K queries/day at an enterprise customer-service division, RouteNLP reduced inference costs by 58% while maintaining 91% response acceptance and reducing p99 latency from 1,847 ms to 387 ms. On a six-task benchmark spanning finance, customer service, and legal domains, the framework achieves 40-85% cost reduction while retaining 96-100% quality on structured tasks and 96-98% on generation tasks, with human evaluation confirming that 74.5% of routed generation outputs match or exceed frontier-model quality.