whatmodel.sh / tts / Trendyol-TTS
Trendyol-TTS
↗ share
+ contribute
Trendyol-TTS is an open-weights text-to-speech model from Trendyol, released June 2026. Its published size is 2.38B parameters. Licence: mit.
Catalogue identity notes
Trendyol's Turkish TTS research model built on openbmb/VoxCPM2 (merged LoRA checkpoint step_0002000; Trendyol/Trendyol-TTS, pinned 66a8018, MIT). The Antalia re-run harness loads VoxCPM.from_pretrained(hf_model_id='Trendyol/Trendyol-TTS') with cfg 2.0 and 16 inference steps; the EMA card lists 'Trendyol-TTS 2.38B'.
Model information sources
Release date ↗ · Weights licence ↗
2.38B as printed in both evaluators' tables (HF safetensors metadata reports 2.29B for the repo's safetensors).
Score within the Vapi Humanness · Bradley–Terry rating, 2026-09-25 snapshot comparison set only.
$ how it compares
The text-to-speech models ranked either side of Trendyol-TTS, out of 148 with a score.
| # | model | score | where |
|---|---|---|---|
| 110 |  Antalia-2 Mini | 40±4 | comparison only |
| 111 |  HierSpeech++ v1.1 | 39±3 | local |
| 112 | Trendyol-TTS | 39±3 | comparison only |
| 113 |  Magpie TTS Multilingual 357M | 39±6 | local |
| 114 |  Zonos v0.1 | 39±4 | api price unknown +local |
see all text-to-speech models · find the ones that fit your machine
$ what it's good at
other tests1 not in a task · 97% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Freya-TR-Eval Turkish TTS intelligibility (WER, faster-whisper large-v3, 8 kHz band) +1 | 1.2% ✎ | ▄··▄··▄···▂█ beats 8/10 | 97% |
$ where to run it
No verified API or download recorded yet · add one
Official weights and documentation: huggingface.co/Trendyol ↗
$ how the score was worked out · Antalia-2 Mini model card (cl… 53%, EMA Lightning model card (can… 43%
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 Antalia-2 Mini model card (cloud0day3/antalia-mini), public Freya-TR-Eval re-run table | 1 | 53% |
| EMA Lightning model card (canberkkkkkk/ema-lightning), Benchmarks: Freya-TR-Eval | 1 | 43% |
| prior · pull toward the middle | 3% | |
| weighted average | = 39.3 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json