Which voice engine should I use?
Use V4 Real for the most natural voice (4 credits per character, Creator plan and above), V3 Expressive for named character voices on the free plan (2 credits), V2 Multilingual for 31 languages including Bulgarian (1.5), and V1 Instant for the cheapest drafts (1).
How do the four engines compare?
- V4 Real: 30 voices, with a style direction and acting tags. 4 credits per character, up to 5,000 characters per generation. Creator and above.
- V3 Expressive: 9 named voices (Atlas, Reid, Sage, Kai, Cole, Ivy, Mira, Elle, Nova) with style direction. 2 credits per character, up to 5,000. Every plan.
- V2 Multilingual: 10 voices (M1 to M5, F1 to F5) in 31 languages. 1.5 credits per character, up to 5,000. Every plan.
- V1 Instant: more than 40 voices in 8 language groups, English first. 1 credit per character, up to 15,000. Every plan.
How do I make a voice sound more expressive?
On V3 and V4 you can write a style direction of up to 300 characters, such as 'warm and slow'. Studio also has delivery presets like calm, excited, whispered and warm and friendly.
On V4 you can put acting tags inline, for example 'That was close <laugh> really close.' Tags include laugh, sigh, gasp, whispers, short pause and long pause. They are never billed, and the other engines strip them out.
Which one should I start with?
- Publishing something people will listen to closely: V4.
- Good quality on the free plan, mostly English: V3. Studio opens on it.
- Text that is not English, on the free plan: V2. See Which languages does Audexum support?
- Long English drafts where cost matters most: V1.
Relacionado