# AI use cases for AuK (Tencent-Hunyuan/AuK) Repo: [Tencent-Hunyuan/AuK](https://github.com/Tencent-Hunyuan/AuK) | Use cases: 5 | Page: https://dedupedai.com/use-cases/tencent-hunyuan-auk/ AuK is an open-source 1.5B speech model that generates and edits speech from natural-language instructions. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Generate speech in a reference speaker's voice: https://dedupedai.com/use-cases/tencent-hunyuan-auk/clone-voice-from-reference-clip/ - Create speech from a voice description alone: https://dedupedai.com/use-cases/tencent-hunyuan-auk/create-speech-from-voice-description/ - Replace or remove words in an existing recording: https://dedupedai.com/use-cases/tencent-hunyuan-auk/edit-words-in-existing-recording/ - Remove noise and reverb from a voice recording: https://dedupedai.com/use-cases/tencent-hunyuan-auk/clean-noisy-speech-recording/ - Turn a plain-language request into a ready-to-run command: https://dedupedai.com/use-cases/tencent-hunyuan-auk/turn-free-form-request-into-auk-command/ ## How to generate speech in a reference speaker's voice with AuK Repo: [Tencent-Hunyuan/AuK](https://github.com/Tencent-Hunyuan/AuK) | Difficulty: Intermediate | Time: 1 hour | Basis: From the README Turn a script into spoken audio that keeps the voice of a short reference recording using zero-shot TTS. **Who it is for:** A content creator or narrator who has a short clip of a voice they are allowed to use and needs a new line read in that voice. **The problem:** Recording every new script line again is slow, and the voice must match the earlier recording. **The outcome:** A WAV file with the target text spoken in the reference voice at the chosen length. ### Steps 1. `git clone https://github.com/Tencent-Hunyuan/AuK` 2. `cd AuK` 3. `uv venv --python 3.10` 4. `source .venv/bin/activate` 5. `uv pip install -e .` 6. `hf download tencent/AuK --local-dir ./ckpts/AuK` 7. `hf download Qwen/Qwen2.5-Omni-3B --local-dir ./ckpts/Qwen2.5-Omni-3B` (inferred, check the README) 8. `auk-infer --audio --instruction "Say the following with the same voice: ''" --output --gen_seconds 6.0` (inferred, check the README) ### Prompt to paste into your agent ```text Using the Tencent-Hunyuan/AuK repo, set up the environment and download the AuK and Qwen2.5-Omni-3B weights, then run auk-infer with and the target text to write . Verify that exists and plays back at about the requested length, and report the path only if the speaker permission for is confirmed. ``` ### Caveat The README gives a duration hint through --gen_seconds, so the output length may need adjusting for your script. Risk, voice cloning: The output imitates the reference speaker, so use only voices you have permission to clone. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/tencent-hunyuan-auk/clone-voice-from-reference-clip/ ## How to create speech from a voice description alone with AuK Repo: [Tencent-Hunyuan/AuK](https://github.com/Tencent-Hunyuan/AuK) | Difficulty: Intermediate | Time: 1 hour | Basis: From the README Generate a spoken line from a text description of the voice, with no reference audio needed. **Who it is for:** A game or app developer who needs a placeholder narrator voice and has no recording to start from. **The problem:** Finding or recording a voice for prototype dialogue takes time before the content can be tested. **The outcome:** A WAV file with the requested text spoken in the described voice. ### Steps 1. `uv venv --python 3.10` 2. `source .venv/bin/activate` 3. `uv pip install -e .` 4. `hf download tencent/AuK --local-dir ./ckpts/AuK` 5. `auk-infer --instruction "" --output --gen_seconds 6.0` (inferred, check the README) ### Prompt to paste into your agent ```text In the Tencent-Hunyuan/AuK repo, use auk-infer with only an --instruction that describes the voice and contains , and write the result to . Verify that is created and that its length is close to the requested seconds. ``` ### Caveat The README lists instruct TTS as a task but does not show its exact command, so the example call here is inferred from the general instruction pattern. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/tencent-hunyuan-auk/create-speech-from-voice-description/ ## How to replace or remove words in an existing recording with AuK Repo: [Tencent-Hunyuan/AuK](https://github.com/Tencent-Hunyuan/AuK) | Difficulty: Intermediate | Time: 1 hour | Basis: From the README Change what is said in a recording by describing the edit in the instruction, keeping the voice and rest of the audio. **Who it is for:** A podcast editor or voice-over producer who needs to fix one phrase in a finished take without re-recording it. **The problem:** Re-recording a single sentence means booking the speaker again and matching the room and tone. **The outcome:** An edited WAV file where the named words are replaced, inserted, or removed. ### Steps 1. `git clone https://github.com/Tencent-Hunyuan/AuK` 2. `cd AuK` 3. `uv venv --python 3.10` 4. `source .venv/bin/activate` 5. `uv pip install -e .` 6. `hf download tencent/AuK --local-dir ./ckpts/AuK` 7. `hf download Qwen/Qwen2.5-Omni-3B --local-dir ./ckpts/Qwen2.5-Omni-3B` (inferred, check the README) 8. `auk-infer --audio --instruction "Replace '' with ''." --output --gen_seconds ` (inferred, check the README) ### Prompt to paste into your agent ```text Using the Tencent-Hunyuan/AuK repo, run auk-infer on with an instruction that replaces with , and write the result to . Verify by listening for the new words and checking that exists with a length close to . ``` ### Caveat The README's content editing example sets a fixed --gen_seconds, so you need to pick a length that matches the edited sentence. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/tencent-hunyuan-auk/edit-words-in-existing-recording/ ## How to remove noise and reverb from a voice recording with AuK Repo: [Tencent-Hunyuan/AuK](https://github.com/Tencent-Hunyuan/AuK) | Difficulty: Intermediate | Time: 1 hour | Basis: From the README Restore a noisy or reverberant speech clip to clearer audio with one instruction and no separate tools. **Who it is for:** A researcher or creator with field recordings or home-studio takes that have background noise or echo. **The problem:** Noisy or echoey recordings are hard to use in a finished project without manual cleanup. **The outcome:** A cleaned WAV file of the speech, with the same length as the input. ### Steps 1. `uv venv --python 3.10` 2. `source .venv/bin/activate` 3. `uv pip install -e .` 4. `hf download tencent/AuK --local-dir ./ckpts/AuK` 5. `hf download Qwen/Qwen2.5-Omni-3B --local-dir ./ckpts/Qwen2.5-Omni-3B` (inferred, check the README) 6. `auk-infer --audio --instruction "Restore this audio to clean speech: keep all speakers, remove noise and reverberation, and keep the same length." --output ` (inferred, check the README) ### Prompt to paste into your agent ```text In the Tencent-Hunyuan/AuK repo, run auk-infer on with an instruction asking for noise and reverberation removal, and save the output as . Verify that exists and has the same duration as . ``` ### Caveat The README's enhancement example uses a Chinese instruction, so an English instruction is untested in the README and may need tuning. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/tencent-hunyuan-auk/clean-noisy-speech-recording/ ## How to turn a plain-language request into a ready-to-run command with AuK Repo: [Tencent-Hunyuan/AuK](https://github.com/Tencent-Hunyuan/AuK) | Difficulty: Advanced | Time: 1.5 hours | Basis: From the README Use Prompt Enhancer to convert a free-form audio request into a complete auk-infer command and manifest. **Who it is for:** A team member who is not comfortable with the CLI flags and wants AuK to choose the task, duration, and preprocessing. **The problem:** Picking the right instruction, target duration, and audio preprocessing for each task is easy to get wrong by hand. **The outcome:** A printed auk-infer command, a detected task and target duration, and a JSON manifest under assets/after_pe/. ### Steps 1. `uv venv --python 3.10` 2. `source .venv/bin/activate` 3. `uv pip install -e ".[gradio]"` 4. `set -a` 5. `source ./.env` 6. `set +a` 7. `python src/auk/infer/pe.py --audio --instruction "" --asr auto` (inferred, check the README) ### Prompt to paste into your agent ```text In the Tencent-Hunyuan/AuK repo, set the OpenAI-compatible LLM variables from .env, then run src/auk/infer/pe.py on with the request . Verify that the printed auk-infer command and the JSON manifest under assets/after_pe/ are created, then run the printed command only after checking its output path. ``` ### Caveat The README says Prompt Enhancer uses the same LLM environment variables described above, but the variable names are not listed in the text provided, so check the repo docs before setup. Risk, credentials: Prompt Enhancer reads OpenAI-compatible LLM settings from .env, so keep that file out of version control and do not paste the key into logs. Dual-use notice: only use this on software, sites and systems you own or are authorized to test. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/tencent-hunyuan-auk/turn-free-form-request-into-auk-command/