Microsoft Research

VALL-E X

Neural codec language model for TTS.

Zero-shot TTSEmotional speech synthesisFew-shot voice cloningOpen source potential
Today's score
83.0
Try VALL-E X

Where it ranks today

Best for / Not great for

Best for
  • Research in TTS
  • Prototyping voice cloning
  • Expressive speech generation
  • Customizable voice applications
Not great for
  • Commercial deployment (currently)
  • Direct music generation
  • Speech-to-text transcription
  • Stable, production-ready API

Why it ranks here

While primarily a research project, VALL-E X is highly influential for its zero-shot and few-shot capabilities in text-to-speech, demonstrating advanced voice cloning and emotional expression that pushes the boundaries of TTS.

30-day trend

Score breakdown

Search trends83
Benchmarks82
Developer buzz85
News mentions83

Pricing

API: $0.00 in · $0.00 out per 1M tokens · Consumer: $0.00/mo

Pricing plans

Popular
Research Code
Access the research code.
Free
  • Zero-shot TTS
  • Few-shot voice cloning
  • Emotional speech
  • Research oriented
View Project
Compare with another modelHow is this score calculated? →Snapshot 2026-10-03