SUTRA are ultrafast, multilingual, online Generative AI models that can operate in 50+ languages with conversation, search, and visual capabilities.
Multilingual
Supports ultrafast text-generation and instruction-following in multiple languages.
Cost-efficient
Energy and cost efficient as well as easier to maintain and add new language capabilities.
Ultrafast
Lightweight MoE architecture and purpose-built tokenizer enables ultra low latency.
Online
Connected and hallucination-free models that provide factual responses in a conversational tone.
50+ languages with
accuracy & efficiency
MMLU scores
Cost-efficient tokenization
for non-English languages
SUTRA's purpose-built tokenizer is 5-8X more cost efficient.
Up-to-date &
hallucination-free
SUTRA-Online are internet connected and hallucination-free models that understand queries, browse the web, and summarize information to provide current answers.
It can answer queries like “Who won the game last night” or “What’s the current stock price” accurately, whereas offline models suffer from knowledge cut-off dates.
Models
SUTRA-Pro
SUTRA-Pro is adept at executing instructions in 50+ languages for conversational use cases and complex tasks. Developed with a rich diversity of instructions from both proprietary and leading open-access datasets, our models excel across a spectrum of multilingual benchmarks, offering unparalleled proficiency in over 50 languages including the nuanced spectrums of Latin, Indic, and Far Eastern languages.
SUTRA-Light
SUTRA-Light is designed for conversation, summarization, and other tasks in 50+ languages. An ideal choice for multilingual conversations at scale, balancing performance, cost, and speed. SUTRA-Light is highly efficient and optimized for ultra-low latency applications.
SUTRA-Online
SUTRA-Online can use knowledge from the internet, and thus leverage the most up-to-date information when forming responses. By understanding queries, browsing the web, and summarizing information, SUTRA-Online can accurately respond to time-sensitive queries, unlocking knowledge beyond its training corpus. This means that SUTRA models can respond to queries like “Who won in yesterday’s India vs England cricket match?” accurately, whereas offline models suffer from knowledge cut-off dates.
Detailed SUTRA model spec and sample code are available.
SUTRA API
available now
SUTRA models are the technology backbone of TWO’s products and services and are available as Model as a Service (MaaS) to other apps and services via usage-based pricing and simple-to-integrate APIs.