Coval
Simulation & evals to ship delightful voice & chat AI agents
Universal-1
Multilingual speech AI model trained on 12.5M hours of data
Try AssemblyAI's most capable and highly trained speech recognition model trained on 12.5M hours of multilingual audio data. Universal-1 achieves best-in-class speech-to-text accuracy, reduces word error rate and hallucinations, and improves timestamps.
Founded on May 21, 2024, Universal-1 has been active for almost 2 years. The product targets the Artificial Intelligence (AI), Developer Tools, and Audio sectors. With 366 votes and 20 comments, it stands out in its category as an up-and-coming startup.
Browse all startups in Artificial Intelligence (AI), Developer Tools, Audio.
Similar startups
Other startups building in the same space
AssemblyAI - Speech-to-Text API
APIs to automatically transcribe and understand audio
DeepGram
Search engine for speech
Vapi
Voice AI agents for Developers
Vapi
Voice AI infrastructure for the internet
Wordware
Your tool for building AI agents with natural language
Lovable
The world's first AI Full Stack Engineer
Wispr Flow
Speak naturally, write perfectly & 3x faster in every app
MGX
The first AI dev team
...
Discussion 0