mlsanityData to 14 Oct 2026

FreyaTTS: A Compact Tokenizer-Free Flow-Matching Transformer for Turkish-First Speech Synthesis

arXiv 2607.09530 ↗First posted 10 Jul 2026cs.CLcited by 0

On the map: Diffusion models → Audio, speech and music

Abstract (arXiv)

We introduce Freya-TTS, a compact, tokenizer-free, Turkish-first text-to-speech model designed for highly reliable and efficient conversational synthesis. Freya-TTS is a 183.2M-parameter non-autoregressive conditional flow-matching Diffusion Transformer (DiT) that operates in the continuous latent space of the frozen AudioVAE2 (16 kHz encode, 48 kHz decode), allowing the model to focus its capacity on text-to-latent mapping while inheriting high-quality 48 kHz reconstruction. We advance the framework along three key dimensions: (1) rule-free end-to-end modeling from a 92-symbol Turkish character vocabulary without a phonemizer, grapheme-to-phoneme frontend, or discrete speech tokenizer, with digit strings expanded to their spoken form at the text frontend; (2) non-autoregressive parallel denoising, which predicts the entire latent sequence simultaneously over a predicted duration; and (3) a production-oriented two-stage post-training recipe consisting of single-speaker voice locking and short-utterance coverage, improving speaker consistency and robustness on short inputs. On the Freya-TR-Eval benchmark, Freya-TTS achieves a band-matched word error rate (WER) of 8.0% and character error rate (CER) of 3.0%, lower error than both larger open systems in its field, XTTS-v2 and F5-TTS, at 40-55% of their parameter count, together with the highest naturalness (MOS) among the compact systems. The model achieves a real-time factor of 0.11 on a consumer GPU (RTX 4090; ~0.14 mean on an H100) and synthesizes in real time on a laptop CPU, making it well suited for resource-constrained edge deployment. We release the model weights, training and inference code, and evaluation benchmark under the Apache-2.0 license.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Audio, speech and music, one of 14 directions in Diffusion models. Builds on 44 of its 45 references that are in the map. First posted on arXiv 10 Jul 2026.

Builds on

44 of 45 references.

Closest ancestors · 10

Same direction, newest first

  1. 2026OmniVoice: Towards Omnilingual Zero-Shot Text-to-Speech with Diffusion Language Models1 Apr
  2. 2025ZipVoice: Fast and High-Quality Zero-Shot Text-to-Speech with Flow Matching16 Jun
  3. 2025CosyVoice 3: Towards In-the-wild Speech Generation via Scaling-up and Post-training23 May
  4. 2025DiTAR: Diffusion Transformer Autoregressive Modeling for Speech Generation6 Feb
  5. 2024F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching9 Oct
  6. 2024NaturalSpeech 3: Zero-Shot Speech Synthesis with Factorized Codec and Diffusion Models5 Mar
  7. 2023Matcha-TTS: A fast TTS architecture with conditional flow matching6 Sep
  8. 2023StyleTTS 2: Towards Human-Level Text-to-Speech through Style Diffusion and Adversarial Training with Large Speech Language Models13 Jun
  9. 2023NaturalSpeech 2: Latent Diffusion Models are Natural and Zero-Shot Speech and Singing Synthesizers18 Apr
  10. 2021Grad-TTS: A Diffusion Probabilistic Model for Text-to-Speech13 May

Other references · 34

  1. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
  2. 2022Scalable Diffusion Models with Transformerscited by 6,054
  3. 2022Flow Matching for Generative Modelingcited by 5,702
  4. 2021RoFormer: Enhanced Transformer with Rotary Position Embeddingcited by 4,703
  5. 2022Robust Speech Recognition via Large-Scale Weak Supervisioncited by 4,057
  6. 2019Common Voice: A Massively-Multilingual Speech Corpuscited by 1,384
  7. 2023Improving and generalizing flow-based generative models with minibatch optimal transportcited by 879
  8. 2017Natural TTS Synthesis by Conditioning WaveNet on Mel Spectrogram Predictionscited by 804
  9. 2023Neural Codec Language Models are Zero-Shot Text to Speech Synthesizerscited by 752
  10. 2020FastSpeech 2: Fast and High-Quality End-to-End Text to Speechcited by 717
  11. 2022High Fidelity Neural Audio Compressioncited by 687
  12. 2024Autoregressive Image Generation without Vector Quantizationcited by 651
22 more
  1. 2020ECAPA-TDNN: Emphasized Channel Attention, Propagation and Aggregation in TDNN Based Speaker Verificationcited by 628
  2. 2022AudioLM: a Language Modeling Approach to Audio Generationcited by 616
  3. 2021Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speechcited by 561
  4. 2022FLEURS: Few-shot Learning Evaluation of Universal Representations of Speechcited by 469
  5. 2023Finite Scalar Quantization: VQ-VAE Made Simplecited by 440
  6. 2023High-Fidelity Audio Compression with Improved RVQGANcited by 431
  7. 2023Scaling Speech Technology to 1,000+ Languagescited by 411
  8. 2023Voicebox: Text-Guided Multilingual Universal Speech Generation at Scalecited by 379
  9. 2024CosyVoice: A Scalable Multilingual Zero-shot Text-to-speech Synthesizer based on Supervised Semantic Tokenscited by 334
  10. 2023The Impact of Positional Encoding on Length Generalization in Transformerscited by 327
  11. 2024Seed-TTS: A Family of High-Quality Versatile Speech Generation Modelscited by 300
  12. 2024Emilia: An Extensive, Multilingual, and Diverse Speech Dataset for Large-Scale Speech Generationcited by 204
  13. 2021SpeechT5: Unified-Modal Encoder-Decoder Pre-Training for Spoken Language Processingcited by 180
  14. 2024XTTS: a Massively Multilingual Zero-Shot Text-to-Speech Modelcited by 179
  15. 2024E2 TTS: Embarrassingly Easy Fully Non-Autoregressive Zero-Shot TTScited by 152
  16. 2026Qwen3-TTS Technical Reportcited by 118
  17. 2025Spark-TTS: An Efficient LLM-Based Text-to-Speech Model with Single-Stream Decoupled Speech Tokenscited by 113
  18. 2024Autoregressive Speech Synthesis without Vector Quantizationcited by 65
  19. 2025CFG-Zero*: Improved Classifier-Free Guidance for Flow Matching Modelscited by 52
  20. 2025MiniCPM4: Ultra-Efficient LLMs on End Devicescited by 52
  21. 2025VoxCPM: Tokenizer-Free TTS for Context-Aware Speech Generation and True-to-Life Voice Cloningcited by 42
  22. 2026VoxCPM2 Technical Reportcited by 17

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025TaDiCodec: Text-aware Diffusion Speech Tokenizer for Speech Language Modeling
  2. 2025MELA-TTS: Joint transformer-diffusion model with representation alignment for speech synthesis
  3. 2026ARCHI-TTS: A flow-matching-based Text-to-Speech Model with Self-supervised Semantic Aligner and Accelerated Inference
  4. 2025OZSpeech: One-step Zero-shot Speech Synthesis with Learned-Prior-Conditioned Flow Matching
  5. 2025M3-TTS: Multi-modal DiT Alignment & Mel-latent for Zero-shot High-fidelity Speech Synthesis
  6. 2024SimpleSpeech 2: Towards Simple and Efficient Text-to-Speech with Flow-based Scalar Latent Transformer Diffusion Models
  7. 2026Fr'echet Distance Loss on Speech Representations for Text-to-Speech Synthesis
  8. 2024On the Semantic Latent Space of Diffusion-Based Text-to-Speech Models