A4F
v2.24.0•

Introducing Audio Generation & Embeddings

Sreejan
SreejanAuthor

This release introduces major new modalities to the A4F platform, enabling developers to work with audio and text embeddings through our unified API.

New Features

2
  • Text-to-Speech (TTS) API: Implemented a new /v1/audio/speech endpoint for generating high-quality audio from text. This includes support for multiple voices and output formats.
  • Embeddings API Support: Added support for text embedding generation through the /v1/embeddings endpoint for a key provider (Provider-3), allowing for the creation of vector representations of text.

Improvements

1
  • Input Validation: Added input length validation for TTS requests, ensuring that prompts adhere to plan-based character limits for better error handling and cost control.

Refactor

1
  • Pricing Calculation: Refactored the pricing mechanism for TTS services to be based on character count rather than token count, providing more accurate and predictable billing.