Docs / Guides
5-Minute Quick-Start
Step 1: Read a Passage the Free Way First
Open Soniva, paste some text into the input box, and press Enter.
By default it uses system built-in voice — the reading engine already on your device. No login, no subscription, no credits deducted, and no usage limit. Confirm that "it can read text aloud" first, then decide whether to upgrade the voice.
Step 2: Hear the Difference of Cloud AI Voices
System built-in voices are decided by your operating system; Soniva can't change them. For a more natural result you need cloud AI voices.
No need to pay first: Alibaba Cloud, Doubao Speech, MiniMax, Tencent Cloud, Baidu, and Microsoft each provide 1 representative voice + 1 fixed script, playable free 5 times per brand per day. ElevenLabs and Cartesia voice lists come with official preview audio that plays immediately.
In TTS Settings (the voice button to the left of the input box), play through each brand and pick one you like.
Step 3: Pick a Brand by Your Needs
There's no "best voice" — only "the voice that fits your text." A few starting points:
| Your need | Try first |
|---|---|
| Chinese reading, balancing quality and price | Alibaba Cloud Bailian · CosyVoice v3-flash (default tier, 16 credits/1,000 chars) |
| Chinese, want it cheaper | Tencent Cloud · Premium Voice or Baidu · Basic Voice Library (both 5 credits/1,000 chars) |
| Chinese, want emotion and character | Doubao Speech · 2.0, or MiniMax |
| Chinese dialects | Doubao Speech · LLM 1.0 (Bay Area Uncle, Adorable Sichuan Girl, Beijing Lad, Guangxi Yuanzhou) |
| English / broadest multilingual coverage | Microsoft Azure · Neural TTS |
| Prioritize English expressiveness, price irrelevant | ElevenLabs (most expensive tier — mind the consumption rate) |
| Very long articles, just want it read through | Doubao Speech · Small Model (~4 credits/1,000 chars, most economical on the site) |
Detailed notes on each are in the Eight Brands at a Glance.
Check the consumption rate before choosing: after picking a brand and model in TTS Settings, "~N credits / 1,000 chars" appears right below the dropdown. Switching model tiers within the same brand can change the price several-fold. Full comparisons in the credit rate table.
Using cloud AI voices requires an active subscription (any of Plus / Pro / Max); see How to Choose a Subscription Tier.
Step 4: These Boundaries Are All You Need
- Per-reading length: 500 chars max per request with credits, 4,000 with your own API key. Anything over auto-segments and reads in sequence; you don't split manually.
- Speech rate: five levels — 0.75x / 1x / 1.25x / 1.5x / 2x.
- Translation quota is separate: pre-render translation doesn't consume credits; the free tier gets 5,000 chars/day.
- Reading your credits: the "Usage" page shows total balance, how much of the grant/top-up remains, and how many credits each reading deducted.
Shortcuts
| Shortcut | Function |
|---|---|
| Enter | Send text and start reading |
| Shift + Enter | New line inside the input box |
Use the bottom play control bar for pause, switching voices, and adjusting the speech rate.
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