# LangSelect: Cost-Aware Target-Language Routing for LLM Code Generation

arXiv: https://arxiv.org/abs/2609.18959 · HTML: https://mlsanity.com/paper/2609.18959/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 18 of its 31 references that are in the map.
- First posted on arXiv 15 Jul 2026.

## Builds on (most cited first, 18 of 18)

- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Program Synthesis with Large Language Models](https://mlsanity.com/paper/2108.07732/) (2021)
- [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023)
- [Neural Machine Translation of Rare Words with Subword Units](https://mlsanity.com/paper/1508.07909/) (2015)
- [SentencePiece: A simple and language independent subword tokenizer and detokenizer for Neural Text Processing](https://mlsanity.com/paper/1808.06226/) (2018)
- [CodeBERT: A Pre-Trained Model for Programming and Natural Languages](https://mlsanity.com/paper/2002.08155/) (2020)
- [Measuring Coding Challenge Competence With APPS](https://mlsanity.com/paper/2105.09938/) (2021)
- [CodeGen: An Open Large Language Model for Code with Multi-Turn Program Synthesis](https://mlsanity.com/paper/2203.13474/) (2022)
- [FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performance](https://mlsanity.com/paper/2305.05176/) (2023)
- [RouteLLM: Learning to Route LLMs with Preference Data](https://mlsanity.com/paper/2406.18665/) (2024)
- [DS-1000: A Natural and Reliable Benchmark for Data Science Code Generation](https://mlsanity.com/paper/2211.11501/) (2022)
- [CodeGeeX: A Pre-Trained Model for Code Generation with Multilingual Benchmarking on HumanEval-X](https://mlsanity.com/paper/2303.17568/) (2023)
- [Chain of Draft: Thinking Faster by Writing Less](https://mlsanity.com/paper/2502.18600/) (2025)
- [Byte Pair Encoding is Suboptimal for Language Model Pretraining](https://mlsanity.com/paper/2004.03720/) (2020)
- [A Comparative Study of Programming Languages in Rosetta Code](https://mlsanity.com/paper/1409.0252/) (2014)
- [Transforming Software Development: Evaluating the Efficiency and Challenges of GitHub Copilot in Real-World Projects](https://mlsanity.com/paper/2406.17910/) (2024)
- [Mix-of-Language-Experts Architecture for Multilingual Programming](https://mlsanity.com/paper/2506.18923/) (2025)
- [Programming Language Confusion: When Code LLMs Can't Keep their Languages Straight](https://mlsanity.com/paper/2503.13620/) (2025)

## Abstract (arXiv)

LLM code-generation systems usually choose a target programming language before decoding and treat that choice as fixed. We show that, for language-flexible programming tasks -- tasks where several target languages are acceptable and checkable by the same tests -- this choice is a measurable cost lever: verified implementations of the same task can differ substantially in generated-token length. We introduce LangSelect, a verification-aware router that selects the target language before generation and falls back when the first attempt fails. To separate offline routing opportunity from end-to-end behavior, we evaluate verified-solution replay, which chooses among already accepted corpus solutions, and live GPT-5 generation, which charges every generation attempt, including failures and fallbacks. On MultiLang-Bench, a 3,000-task, 8-language verified corpus, replay shows substantial language-routing headroom. In live evaluation on 450 held-out tasks, a train-split Domain heuristic baseline reduces harness-proxy tokens, which include wrapper and entrypoint overhead, by 50.3% at 92.9% pass after fallback, while a learned CodeBERT+metadata selector reaches the highest pass after fallback, 93.8%, with a 3.7% token increase. These results show that output-language routing can define a practical cost-correctness frontier for unit-test-verifiable code generation.