Hugging Face launches Open TTS Leaderboard for multilingual text-to-speech evaluation
Hugging Face introduced the Open TTS Leaderboard on September 30, 2026, to address gaps in text-to-speech (TTS) model evaluation. Existing arena-style leaderboards rely on human preference scores but struggle with scalability and open-source representation, listing only 16 of 92 models as open-weights. The new leaderboard uses objective metrics—word/character error rates (WER/CER), inference…
Key points
- Open TTS Leaderboard uses objective metrics (WER, RTFx, SIM) instead of human preference for faster, scalable evaluation
- Only 16 of 92 TTS models on arena-style leaderboards are open-weights, per September 30 data
- Models like `hexgrad/Kokoro-82M` rank top in English WER, but multilingual performance varies by language
The leaderboard prioritizes open-source models and multilingual support, offering a 'Listen' tab for direct audio comparisons and a 'Streaming' tab for latency benchmarks. Models like hexgrad/Kokoro-82M and fishaudio/s2-pro lead in English WER, while k2-fsa/OmniVoice excels in multilingual performance. Hugging Face plans to open-source evaluation scripts and invites community feedback to refine future iterations.
Open TTS Leaderboard: Scalable Evaluation for Multilingual Text-to-Speech and Voice Cloning
Hugging Face · 30 September 2026
Loading the full article…
This text was published by Hugging Face and written by Eric Bezzam. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Research
All →- Nvidia pledges $250M in compute for UK AI agent safety research · 8 src
- OpenAI offers U.S. university labs free ChatGPT Business seats with Rosalind access · 2 src
- Anthropic reports small AI models may self-replicate on IoT devices within months · 29 src
- Researchers propose LLM-referenced C-test for evaluating supra-intelligence · 8 src
- Researchers study permutation diffusion LLMs for recommendation reranking · 1 src
Comments
via GitHub Discussions