Meta
Meta-VoiceSynth-2023
audio · Release Jul 25, 2023 · License n/a
A neural voice synthesis model capable of producing realistic human speech with emotional intonations.
Modalities
What goes in and what comes out.
Inputs
text
Outputs
audio
Capabilities
text-to-speech, voice customization, emotion synthesis, multi-language support
Benchmarks snapshot
Structured JSON for reproducible comparisons.
{
"naturalness_score": 9.5
}Related on GenAIWiki
Same provider, tooling that cites the model, or prompts tuned for it.
Meta
LLaMA 3 70B
LLaMA 3 70B features 70 billion parameters and a context window of 32k tokens, optimized for high-performance text generation and understanding across diverse tasks.
Meta
LLaMA 3 8B
LLaMA 3 8B is a compact model with 8 billion parameters, designed for efficient text generation and understanding with a context window of 8k tokens.
Meta
Llama 3.1 405B Instruct
Large open-weights instruct model competitive on reasoning and coding benchmarks with permissive licensing for customization.