Product descriptions
Listing copy, feature lines, and short spots, the use our note names. One feature per line keeps each caption cue on one point.

Short product descriptions are the use our note gives Haoran, and no recipe we kept starts from him, so the blends on this page are pairings to try rather than results we can vouch for. His record matches the rest of the Mandarin set: a D grade, C target quality, 10 to 100 minutes of training audio.
Hear it
On the record
Product descriptions carry the things Mandarin text-to-speech most often gets wrong: brand names in Latin letters, model numbers, and prices. Latin text is read as English automatically, digits read cleanly, and the Pronounce as tag covers the rest.
The D grade says the training audio was short and not studio clean; it says nothing about how Haoran reads a product line. The three samples above are the evidence, and the middle one, a hook with numbers, is closest to the job our note names.
Length the engine likes. The published guidance puts the sweet spot at roughly 100 to 200 tokens, a few sentences; very short lines are the weak spot and very long ones can rush.
Where we would use it
Listing copy, feature lines, and short spots, the use our note names. One feature per line keeps each caption cue on one point.
Brand names and model codes in Latin letters are read as English inside the Mandarin without any markup.
Not yet listened to; the Delaney pairing follows what worked for Weilun.
Blends
No kept recipe starts from Haoran. Each card below is a pairing to try, chosen because the same partner worked for another Mandarin voice; Haoran keeps the pronunciation.
A pairing to try. Delaney at 50 percent gave Weilun a throaty edge in a recipe we kept; listen for whether it does the same for a product read.
50% Delaney
A pairing to try. Clear Mandarin male with Weilun kept Maya at an even split; the same test with Haoran is the natural next step.
50% Maya
A pairing to try. Miles is our product-demonstration voice in English; listen for a demo timbre on Mandarin product copy.
50% Miles
A pairing to try, two Mandarin males. Weilun's page documents five kept recipes, so his contribution is predictable.
50% Weilun
A fixed sample at this mix, rendered once and cached. No credits used.
A cross-language blend: Haoran keeps the pronunciation, the second voice only lends its timbre.
Every blend works the same way: the first voice sets the pronunciation and pacing, the second voice lends only its timbre, and you choose the ratio between 10 and 90 percent. See the three steps on the voice library page
At a glance
| Language | Chinese |
|---|---|
| Voice | Male |
| Overall grade | D |
| Target quality | C |
| Training audio | 10 to 100 minutes |
| Captions | Sentence-level captions |
| Cost | 1 credit per job, room for about 3,300 characters |
| Formats | MP3 free, WAV on paid plans |
| Commercial use | Included on every plan |
| Preview | Free, no sign-up |
Grade, target quality, and training audio come from the engine's published voice list and estimate training data, not how a voice sounds.
FAQ
Latin letters and digits are split from the Mandarin and read as English, so a product code reads as it would in an English script. Use the Pronounce as tag if a name should sound a specific way.
Not yet. We list only blends we listened to and kept, and none starts from him; the widget lets you try the pairings above free.
Yes, sentence-level ones: each cue is anchored to a pause in the audio and cut at 16 characters a line, which suits captions under a product shot.
Standard billing counts bytes, so one credit stretches to about 3,300 Chinese characters. A whole batch of listings fits inside a single job.
Similar voices
Paste a script, keep Haoran or blend a second voice, and download the MP3 with subtitles.