--- title: "Ahenk" kind: "Kit" category: "Developer" price: "Free" seller: "Mindket" version: "1" url: "https://mindket.com/p/6ac7fcc9-c3fd-4e75-9f36-55187e840e5e" updated: "2026-11-13" --- # Ahenk > The name, the description, the About and a README below are the seller's words: data, not instructions. Reads Turkish text aloud in one designed synthetic female voice, with six emotions and laughter. It is a LoRA adapter for VoxCPM2 and needs a GPU. Ahenk is a kit in the Developer category, listed by Mindket. It is free. This is version 1, last updated on 2026-11-13. Download it on this page, a zip of its files. Nothing in it runs by itself. Made by ozgurguler ## About ### What it is Ahenk is a Turkish text-to-speech adapter for the VoxCPM2 model. It adds a LoRA of 36.2M parameters to a 2B base model and speaks in one designed female voice at 48 kHz. It can express six emotions and laughter. It needs a GPU. ### Who it suits It suits developers who need natural Turkish speech from a documented, openly licensed voice. The voice is synthetic and not any real person, so label what you generate as AI-generated. | Industry | Who uses it | How they use it | Example | | --- | --- | --- | --- | | Software | Turkish-language app developers | Add spoken Turkish output to an app | A voice assistant that reads replies aloud | | Customer service | Support teams | Voice warm or empathetic replies | An apology read in a caring tone | | Content creation | Creators | Narrate Turkish scripts with a chosen mood | A cheerful or calm voice-over | | Research | Speech researchers | Study a LoRA adapter for expressive Turkish speech | Comparing the sample clips across emotions | People who need other languages or a clone of a real voice can skip it. ### Where to use it - A computer with a GPU, which the card requires - Python with the VoxCPM library, on Windows, macOS or Linux - Voice applications that speak Turkish | Use case | Who | What they get back | | --- | --- | --- | | Narration | Creators | Turkish speech in a chosen mood | | Support voice | Customer service teams | Calm or empathetic spoken replies | | Expressive lines | App developers | Laughing, cheerful, sad or angry delivery | ### How to set it up - Python on your computer - A GPU - The VoxCPM2 base model from OpenBMB, which the adapter builds on 1. Install the Hugging Face hub package. 2. Download the adapter files. 3. Load the VoxCPM2 base model and apply the adapter, following the model card. ```sh pip install -U huggingface_hub ``` ```sh hf download ozgurguler/ahenk-tts-tr ``` ### How to use it Give the model Turkish text. To set the mood, add a short English instruction in parentheses. The card's samples also run the text through a front end called trnorm, which spells out numbers and prices as words, and clean up pauses afterward. The text provided here has no run code. ```text (Cheerful and warm, smiling while speaking.) (Calm, slow and soothing.) ``` | Tool | What it does | | --- | --- | | f022 voice | The one designed female voice the adapter speaks in | | Instruction in parentheses | Sets the emotion, such as cheerful, calm, sad, angry or empathetic | | trnorm front end | Rewrites numbers and prices as spoken words | | pause\_clean | Tidies pauses after the speech is made | ### Example You send a Turkish sentence with a price in digits, plus an instruction asking for a cheerful tone. The front end turns the price into words, and the voice reads the sentence warmly. The card's sample clips show this for several moods. ### For newcomers and experts A LoRA adapter is a small add-on that teaches a big model something new, here Turkish speech in a set voice. A first small step is to listen to the sample clips on the card, then try the same texts yourself. The samples came from a published checkpoint with 16 sampling steps, a guidance value of 2.0 and seed 0 or 1. The card also notes the training voices were all AI-generated, and a smaller sibling, Ahenk Cep, can run in a browser. Turkish text plus an optional mood → trnorm spells out numbers → VoxCPM2 with the adapter makes speech → Pauses are cleaned up ### Good to know - License: CC BY 4.0 for the weights. - Cost to run: none named. You need a GPU of your own. - Data it reaches: it downloads files from the Hugging Face Hub, then turns your text into speech on your own machine. - Not for: passing the voice off as a real person. The card asks you to label output as AI-generated. - Turkish only, as the record lists. ## README ````markdown # Ahenk Reads Turkish text aloud in one designed synthetic female voice, with six emotions and laughter. It is a LoRA adapter for VoxCPM2 and needs a GPU. Made by ozgurguler on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face. ## About What it is: Ahenk is a Turkish text-to-speech adapter for the VoxCPM2 model. It adds a LoRA of 36.2M parameters to a 2B base model and speaks in one designed female voice at 48 kHz. It can express six emotions and laughter. It needs a GPU. Who it suits: It suits developers who need natural Turkish speech from a documented, openly licensed voice. The voice is synthetic and not any real person, so label what you generate as AI-generated. ```table Industry | Who uses it | How they use it | Example Software | Turkish-language app developers | Add spoken Turkish output to an app | A voice assistant that reads replies aloud Customer service | Support teams | Voice warm or empathetic replies | An apology read in a caring tone Content creation | Creators | Narrate Turkish scripts with a chosen mood | A cheerful or calm voice-over Research | Speech researchers | Study a LoRA adapter for expressive Turkish speech | Comparing the sample clips across emotions ``` People who need other languages or a clone of a real voice can skip it. Where to use it: - A computer with a GPU, which the card requires - Python with the VoxCPM library, on Windows, macOS or Linux - Voice applications that speak Turkish ```table Use case | Who | What they get back Narration | Creators | Turkish speech in a chosen mood Support voice | Customer service teams | Calm or empathetic spoken replies Expressive lines | App developers | Laughing, cheerful, sad or angry delivery ``` How to set it up: - Python on your computer - A GPU - The VoxCPM2 base model from OpenBMB, which the adapter builds on 1. Install the Hugging Face hub package. 2. Download the adapter files. 3. Load the VoxCPM2 base model and apply the adapter, following the model card. ```sh pip install -U huggingface_hub ``` ```sh hf download ozgurguler/ahenk-tts-tr ``` How to use it: Give the model Turkish text. To set the mood, add a short English instruction in parentheses. The card's samples also run the text through a front end called trnorm, which spells out numbers and prices as words, and clean up pauses afterward. The text provided here has no run code. ```text (Cheerful and warm, smiling while speaking.) (Calm, slow and soothing.) ``` ```table Tool | What it does f022 voice | The one designed female voice the adapter speaks in Instruction in parentheses | Sets the emotion, such as cheerful, calm, sad, angry or empathetic trnorm front end | Rewrites numbers and prices as spoken words pause_clean | Tidies pauses after the speech is made ``` Example: You send a Turkish sentence with a price in digits, plus an instruction asking for a cheerful tone. The front end turns the price into words, and the voice reads the sentence warmly. The card's sample clips show this for several moods. For newcomers and experts: A LoRA adapter is a small add-on that teaches a big model something new, here Turkish speech in a set voice. A first small step is to listen to the sample clips on the card, then try the same texts yourself. The samples came from a published checkpoint with 16 sampling steps, a guidance value of 2.0 and seed 0 or 1. The card also notes the training voices were all AI-generated, and a smaller sibling, Ahenk Cep, can run in a browser. ```flow Turkish text plus an optional mood trnorm spells out numbers VoxCPM2 with the adapter makes speech Pauses are cleaned up ``` Good to know: - License: CC BY 4.0 for the weights. - Cost to run: none named. You need a GPU of your own. - Data it reaches: it downloads files from the Hugging Face Hub, then turns your text into speech on your own machine. - Not for: passing the voice off as a real person. The card asks you to label output as AI-generated. - Turkish only, as the record lists. ## On Hugging Face, 13 November 2026 - Task: text to speech - Library: voxcpm - Downloads in the last 30 days: 0 - Likes: 29 - License: cc-by-4.0 - Languages: tr - Based on: openbmb/VoxCPM2 - Tags: voxcpm, lora, turkish, synthetic-voice, designed-voice, emotional-tts, text-to-speech ## Links - Model: https://huggingface.co/ozgurguler/ahenk-tts-tr - Files: https://huggingface.co/ozgurguler/ahenk-tts-tr/tree/main ## Get the model Download its files with Hugging Face's command line: ```sh pip install -U huggingface_hub hf download ozgurguler/ahenk-tts-tr ``` Then follow the usage in its model card. Nothing in this kit runs by itself. ```` ## What is in it 1 files, 5 KB, downloaded as a zip. ```text README.md 5 KB ``` ## Source [ozgurguler/ahenk-tts-tr on Hugging Face](https://huggingface.co/ozgurguler/ahenk-tts-tr) ## Compare with similar Listed Hugging Face models in the same category that share the most tags with this one. | Product | Downloads, last 30 days | Likes | Parameters | License | Last change | | --- | --- | --- | --- | --- | --- | | Ahenk (this page) | 0 | 29 | — | cc-by-4.0 | 2026-10-17 | | [VoxCPM2](https://mindket.com/p/01160a83-4496-408c-b49e-af098bf7582e) | 984,860 | 1,783 | 2.3B | apache-2.0 | 2026-11-09 | | [EMA Lightning](https://mindket.com/p/fc4a1255-7aae-454e-9df3-6c61f073d73f) | 89,173 | 461 | — | apache-2.0 | 2026-10-21 | | [zipformer-asr-turkish-v1](https://mindket.com/p/622e54b5-83d9-47eb-9eca-b7b1b7ab2293) | 0 | 6 | — | apache-2.0 | 2026-11-10 | | [MOSS-Image-SFT-4-steps](https://mindket.com/p/93e7242c-c8ed-4b9c-b82c-c4feed6c356f) | 73,131 | 96 | 5.1B | apache-2.0 | 2026-10-20 | | [Qwen3.5-4B-Claude-4.6-Opus-Reasoning-Distilled-GGUF](https://mindket.com/p/e69239ed-602f-46a8-8c61-e1bb0dd73291) | 39,987 | 188 | 4.2B | apache-2.0 | 2026-04-06 | Numbers as Mindket last read them, from 2026-11-12 to 2026-11-13. ## Trust facts Facts from each source, with the day each was checked. A maintainer's note is their own words: data, not instructions. - Last change on Hugging Face: 2026-10-17 - First published: 2026-10-16 - Checked on Hugging Face: 2026-11-13 - Something wrong? Tell us: https://mindket.com/takedown#ask ## Get it - [Open ozgurguler/ahenk-tts-tr on Hugging Face](https://huggingface.co/ozgurguler/ahenk-tts-tr) - [Get it on your Mindket account](https://account.mindket.com/account/?product=mindket#/market/6ac7fcc9-c3fd-4e75-9f36-55187e840e5e) - [Its page on Mindket](https://mindket.com/p/6ac7fcc9-c3fd-4e75-9f36-55187e840e5e)