Docs / Providers
Tencent Cloud Speech Synthesis
In one sentence: the brand with the widest price gap within itself — its cheapest tier is the site's lowest price, and its most expensive tier is 21× that. Picking the wrong tier costs more than picking the wrong brand.
In one sentence: the brand with the widest price gap within itself — its cheapest tier is the site's lowest price, and its most expensive tier is 21× that. Picking the wrong tier costs more than picking the wrong brand.
What This Is
Tencent Cloud's speech-synthesis service, with stable Chinese voice quality. Soniva uses its short-text synthesis interface.
In Soniva Tencent Cloud has only one "model" option (standard) — the three price tiers are distinguished by the voices themselves: which voice you pick decides how many credits get deducted. This differs from other brands, so make sure you understand it.
The Three Tiers & Pricing
| Tier | Voice ID prefix | Official upstream price | Credits / 1,000 chars | Gap |
|---|---|---|---|---|
| Premium Voice | 101… | ¥0.3 / 10,000 chars | 5 | Baseline |
| LLM Voice / Multi-Emotion LLM | 501… 601… | ¥0.55–1.2 / 10,000 chars | 19 | 3.8× |
| Super-Natural LLM Voice | 502… 602… 603… | ¥4.9–6.5 / 10,000 chars | 104 | 21× |
Price source: https://cloud.tencent.com/document/product/1073/34112 (verified 2026-08-05)
Premium Voice (5 credits/1,000 chars) ties with Baidu's Basic Voice Library as the site's cheapest. If your need is "read long articles aloud and the quality just needs to be decent," this tier is unbeatable value — with the same credits, Premium Voice reads 21× more characters than the Super-Natural tier.
How to Tell Which Tier You've Selected
After picking a voice in TTS Settings, "~N credits / 1,000 chars" appears directly below the dropdown. Seeing 5 means Premium, 19 means LLM tier, 104 means Super-Natural.
The group names in the voice list also say so: "Super-Natural LLM," "LLM Voice," "Premium Voice."
What the Three Tiers Actually Differ On
- Premium Voice: traditional TTS technology — steady intonation, accurate pronunciation, easy to listen to for a long time. The downside is flatter emotion.
- LLM Voice: more natural tone with some emotional variation; suits content with dialogue and tonal shifts.
- Super-Natural LLM Voice: close to real human recordings, able to express subtle pauses and emotion. Quality is genuinely good, but the price is in the same league as ElevenLabs — recommended only for short content.
One Practical Tip
Tencent Cloud's voice list defaults to putting the Super-Natural tier first
(502007 Zhixiaohu is the first, and the free-preview representative voice). So if you just tap
the first voice and start reading long text, you'll be charged at the most expensive tier.
Before reading long text, scroll down and find a 101-prefixed Premium voice.
Free Quota
Tencent Cloud's upstream new-account free quota: Premium Voice 8M characters (valid 3 months), LLM Voice 100k characters, Super-Natural 20k characters.
This is Soniva's own account's quota, not yours — it's listed here only to explain where the rate table figures come from. Your usage in Soniva is always billed in credits.
Related Documents
Was this article helpful?
Anonymous feedback is enabled when you allow analytics and the service is available. You can also report an issue by email.
The email includes this article. You can also copy the details and send them to soniva.cde@gmail.com