Hear an actual result
Original transcript: Narrivela turns your words into clear, natural speech.
Generated on October 13, 2026, using Qwen Audio 3.0 TTS Plus (qwen-audio-3.0-tts-plus) in Beijing, with its default female voice longanlingxin, ordinary narration, and no style prompt. The script is original demonstration material. The 24 kHz PCM audio returned by the API is unchanged; only the WAV header length fields were normalized for playback.
This short sample demonstrates one female voice. Long-form narration and the male voice need their own listening checks. Check current service and plugin availability separately.
Download the English WAV · Hear the Mandarin Chinese sampleEarlier engine archive
Hear the older Chinese Cherry recording
Original Chinese transcript: 你好，这是 VoxKit 的语音示例。把文字变成声音，让创作继续。
This archival recording was made under the previous product name, VoxKit, during a Qwen speech test on October 3, 2026. The spoken transcript retains that previous name; it has not been regenerated for Narrivela. It uses the Cherry voice, Chinese (zh-CN), and the original style prompt “自然、清晰、亲切” (natural, clear, and friendly).
This is an older engine's sample, not a demo of the engine selected for Narrivela's first public release. Cherry and style prompts are not options in that release. This recording is not an English voice sample. See the first-release voice and language scope before choosing settings.
Download the sample MP3Copy a prompt
The prompts below use different example scripts from the fixed audio samples. The first public release is limited to ordinary Mandarin Chinese and English narration with two voices. Prefilling the studio does not submit a task or incur speech costs.
Use Narrivela to generate an English MP3 of this narration, using an available voice at speed 1: Today, we start with a small question and look a little closer at the things we encounter every day.
Use Narrivela to submit these two English scripts as one batch. Produce WAV files and preserve their order. Script 1: Start by observing what happens, then ask a question. Script 2: Use an example to explain a possible cause. When the task is complete, provide the individual audio files and a ZIP download.
Check the status of my previous Narrivela speech task. If any items failed and remain unfinished, explain the error first. Once the cause is resolved, resume the unfinished parts of the original task and preserve the completed audio.
What to listen for
Check names, numbers, pauses between sentences, item order, and whether the audio is complete. Test the chosen voice and expression with your own script before committing to a larger batch.