Large Language Model (LLM)
Indic-Bias is a benchmark to evaluate fairness across 85 Indian identity groups focusing on bias and stereotypes.
One API
300+ AI Models
Uptime 99%
Integrate AI features via our secure API
TPD Models
LLaMA Model
LLaMA is an efficient large language model optimized for high-quality text generation.
Retrieval-Augmented Generation (RAG)
We evaluate retrieval-augmented generation across 8 Indian languages with new metrics for retrieval accuracy, generation quality and factuality.
Text-to-Speech (TTS)
We evaluate text-to-speech models across multiple languages focusing on naturalness, intelligibility, and expressive control.
Speech-to-Speech Translation (STS)
STS directly translates spoken language into another spoken language, combining ASR, MT, and TTS in a single system.
Automatic Speech Recognition (ASR)
ASR enables machines to transcribe spoken language into text with high accuracy across multiple domains and accents.
