Fish S2.1 Pro
by Fish Audio
Fish Audio's latest multilingual voice — any reference voice, delivery notes in the script, minimum charge only.
Fish S2.1 Pro is Fish Audio's current-generation speech model, strong in Chinese and covering more than 80 languages. Instead of a fixed voice list it takes any voice id from Fish Audio Discovery — including a voice you have cloned — and you can write delivery directions straight into the script: pauses, emphasis, tone.
In Zalify it is served with no per-character charge, so every render books only the $0.01 minimum. That makes it the obvious pick for long scripts and for markets where an OpenAI voice sounds foreign.
Fish S2.1 Pro pricing on Zalify
Fish S2.1 Pro is billed per 1,000 characters of the script you submit. The voice you pick doesn't change the price, and every render books a minimum of $0.01.
| Rate | Price per 1,000 characters |
|---|---|
| Price per 1,000 characters | Minimum charge only |
Every voiceover books at least $0.01. Models served free through the gateway are billed at that minimum only.
What a typical render costs
300-character voiceover
$0.01
1,000-character voiceover
$0.01
3,000-character voiceover
$0.01
Specifications
Provider
Fish Audio
Voices
Any Fish Audio voice id, incl. cloned voices
Languages
80+
Inline delivery directions
Yes
Where Fish S2.1 Pro shines
80+ languages with strong Chinese
Any Fish Audio voice id, including cloned voices
Delivery directions written inline in the script
Billed at the $0.01 minimum regardless of length
Best for
Frequently asked questions
Fish S2.1 Pro is served at no per-character charge, so every voiceover books only the $0.01 minimum regardless of script length. The price is deducted from your workspace usage balance in US dollars; there is no per-model subscription.
No. Every Zalify workspace with usage balance can generate with Fish S2.1 Pro in Zalify Canvas. Paid plans include a monthly usage allowance; you can also add balance at any time, and purchased balance never expires.
Yes. The generate button in Zalify Canvas shows an estimate for the exact model, resolution and length you selected — for example "Generate video · est. $0.24" — before you commit.
No. You are only charged for renders the provider actually completes. If a generation errors out, nothing is deducted from your balance.
Every model in Zalify Canvas is priced on the same USD basis, so you can compare directly on the AI models overview. Cheaper tiers are ideal for iterating on a concept; switch to the model or resolution you want for the final render.
Other voice models
OpenAI TTS Standard
From $0.0225 per 1,000 characters
OpenAI TTS HD
From $0.045 per 1,000 characters
Grok TTS
From $0.0225 per 1,000 characters
Fish S2 Pro
From $0.01 per 1,000 characters
Try Fish S2.1 Pro in Zalify Canvas
Start free, add balance when you need it, and generate with every model from one canvas — no per-model subscriptions.