# LLM-Powered Grapheme-to-Phoneme Conversion: Benchmark and Case Study

arXiv: https://arxiv.org/abs/2409.08554 · HTML: https://mlsanity.com/paper/2409.08554/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Builds on 11 of its 32 references that are in the map.
- First posted on arXiv 13 Sep 2024.

## Builds on (most cited first, 11 of 11)

- [Qwen Technical Report](https://mlsanity.com/paper/2309.16609/) (2023)
- [Mistral 7B](https://mlsanity.com/paper/2310.06825/) (2023)
- [Gemma 2: Improving Open Language Models at a Practical Size](https://mlsanity.com/paper/2408.00118/) (2024)
- [Mixtral of Experts](https://mlsanity.com/paper/2401.04088/) (2024)
- [Common Voice: A Massively-Multilingual Speech Corpus](https://mlsanity.com/paper/1912.06670/) (2019)
- [FastSpeech 2: Fast and High-Quality End-to-End Text to Speech](https://mlsanity.com/paper/2006.04558/) (2020)
- [ByT5 model for massively multilingual grapheme-to-phoneme conversion](https://mlsanity.com/paper/2204.03067/) (2022)
- [Transformer based Grapheme-to-Phoneme Conversion](https://mlsanity.com/paper/2004.06338/) (2020)
- [ManaTTS Persian: a recipe for creating TTS datasets for lower resource languages](https://mlsanity.com/paper/2409.07259/) (2024)
- [LiteG2P: A fast, light and high accuracy model for grapheme-to-phoneme conversion](https://mlsanity.com/paper/2303.01086/) (2023)
- [Spoken Language Intelligence of Large Language Models for Language Learning](https://mlsanity.com/paper/2308.14536/) (2023)

## Built on by (most cited first, 9 of 9)

- [ShieldedCode: Learning Robust Representations for Virtual Machine Protected Code](https://mlsanity.com/paper/2601.20679/) (2026)
- [PolyIPA -- Multilingual Phoneme-to-Grapheme Conversion Model](https://mlsanity.com/paper/2412.09102/) (2024)
- [LLM-based Generative Error Correction for Rare Words with Synthetic Data and Phonetic Context](https://mlsanity.com/paper/2505.17410/) (2025)
- [Benchmarking Large Language Models for Grapheme-to-Phoneme Conversion: A Japanese Case Study](https://mlsanity.com/paper/2606.22009/) (2026)
- [GraphemeAug: A Systematic Approach to Synthesized Hard Negative Keyword Spotting Examples](https://mlsanity.com/paper/2505.14814/) (2025)
- [P-CoT: A Pedagogically-motivated Participatory Chain-of-Thought Prompting for Phonological Reasoning in LLMs](https://mlsanity.com/paper/2507.16656/) (2025)
- [OLaPh: Optimal Language Phonemizer](https://mlsanity.com/paper/2509.20086/) (2025)
- [Echoes Across Centuries: Phonetic Signatures of Persian Poets](https://mlsanity.com/paper/2603.14443/) (2026)
- [Dictionary-Constrained Grapheme-to-Phoneme for Unsegmented Languages from LLM-Annotated Data](https://mlsanity.com/paper/2609.19805/) (2026)

## Abstract (arXiv)

Grapheme-to-phoneme (G2P) conversion is critical in speech processing, particularly for applications like speech synthesis. G2P systems must possess linguistic understanding and contextual awareness of languages with polyphone words and context-dependent phonemes. Large language models (LLMs) have recently demonstrated significant potential in various language tasks, suggesting that their phonetic knowledge could be leveraged for G2P. In this paper, we evaluate the performance of LLMs in G2P conversion and introduce prompting and post-processing methods that enhance LLM outputs without additional training or labeled data. We also present a benchmarking dataset designed to assess G2P performance on sentence-level phonetic challenges of the Persian language. Our results show that by applying the proposed methods, LLMs can outperform traditional G2P tools, even in an underrepresented language like Persian, highlighting the potential of developing LLM-aided G2P systems.