So many LLMs, so many experts! LLM/Model-Routing holds the promise to combine all of them and answer queries by finding the most compatible, affordable and brisk expert.
Below, we provide a curated list of research papers and studies discussing LLM routing techniques, as well as projects and companies offering LLM and model routing solutions.
Note: if a paper has both an arXiv and a conference/journal proceedings version, then we only show its proceedings version.
- Date: refers to the date of the associated publication (arXiv/proceedings).
- Affiliation: refers to the authors' company, university and/or research labs affiliations.
- Paper: provides the link to the proceedings or arXiv paper version.
- Citations: the number of citations for each paper and the published conference.
- GitHub: paper's repository.
| Date | Affiliation | Paper | Citations | GitHub |
|---|---|---|---|---|
| 05-2023 | Stanford University | Frugalgpt: How to use large language models while reducing cost and improving performance |  |
|
| 09-2023 | Broad Institute, MIT (CSAIL), MIT-IBM Watson AI Lab, University of Michigan | Large language model routing with benchmark datasets |  |
|
| 03-2024 | Martian, UC Berkeley, UC San Diego | ROUTERBENCH: A Benchmark for Multi-LLM Routing System |  |
https://github.com/withmartian/routerbench |
| 05-2024 | MBZUAI | Harnessing the Power of Multiple Minds: Lessons Learned from LLM Routing |  |
https://github.com/kvadityasrivatsa/llm-routing |
| 06-2024 | University of British Columbia, Microsoft, Hippocratic AI | Hybrid LLM: Cost-Efficient and Quality-Aware Query Routing |  |
|
| 07-2024 | UC Berkeley, Anyscale, Canva | RouteLLM: Learning to Route LLMs with Preference Data |  |
https://github.com/lm-sys/RouteLLM |
| 08-2024 | TensorOpera | PolyRouter: A Multi-LLM Querying System |  |
(by alphabetic order)
- Hannibal046/Awesome-LLM on Large Language Models
- KennethanCeyer/awesome-llm on Large Language Models