Sonix
Audio & Voice Translation & NLP
Sonix

Automatic transcription service converting audio/video to text with high accuracy.

Direct link

Sonix uses advanced speech recognition to create searchable transcripts from recordings. Supports speaker identification, timestamps and multiple export formats. The web-based editor allows easy corrections and collaboration. Ideal for journalists, researchers, podcasters and legal professionals needing accurate documentation of spoken content. Integrates with Zoom and other video platforms.

Relevant Sites

Kimi-Audio

Kimi-Audio: A universal open-source audio foundation model handling ASR, AQA, AAC & more. Pre-trained on 13M hours for SOTA performance. Features hybrid architecture & low-latency inference.
Listnr

Versatile AI voice generator with 1000+ voices in 142 languages, including personal voice cloning for unique content creation.
NVIDIA Parakeet-v2

Parakeet-tdt-0.6b-v2: A 600M-parameter ASR model for accurate English transcription with punctuation, capitalization & timestamp prediction. Handles 24-min audio efficiently.
OpenAI Whisper

Whisper is a versatile AI speech recognition model for multilingual transcription, translation, and language ID. Trained on diverse audio data for accurate, general-purpose speech processing.
Replica Studios

AI voice acting platform offering 200+ character voices for game development, animation, and interactive storytelling.
ACE-Step

ACE-Step is an open-source music generation model combining speed, coherence, and control, generating 4-minute tracks in 20 seconds with fine-grained detail.