Different Types of SUTRA Features
This guide outlines the different SUTRA models and their unique features, designed for diverse applications like conversation, reasoning, quantitative analysis, and voice interaction. Each model leverages TWO AI’s dual-transformer architecture for scalability and efficiency.
🌟 SUTRA Model Types
SUTRA-V2
- Purpose: Multilingual conversational AI for chat, search, and instruction tasks.
- Features: Supports 50+ languages, image processing, streaming mode, and OpenAI-compatible API. Outperforms GPT-4 in languages like Gujarati and Tamil.
- Use Cases: Global chatbots, content creation, and cross-language applications.
SUTRA-R0
- Purpose: Advanced reasoning for complex decision-making and multi-step problem-solving.
- Features: Logical reasoning layer, high Multilingual MMLU scores (e.g., Hindi: 81.44), efficient resource usage. Outperforms DeepSeek-R1-32B in Indic languages.
- Use Cases: Financial risk assessment, healthcare diagnostics, customer service automation.
SUTRA-Q0
- Purpose: Time-series quantitative analysis for forecasting and analytics.
- Features: Dense architecture optimized for dynamic data, supports trend analysis and anomaly detection.
- Use Cases: Market forecasting, logistics optimization, healthcare analytics.
SUTRA-Dual2L
- Purpose: Enhanced multilingual model built on Llama 3.1.
- Features: 5-10% improvement in multilingual performance for 70B model, supports Korean, Gujarati, Tamil, and more, with a 256K vocabulary tokenizer.
- Use Cases: Enterprise applications requiring scalable multilingual solutions.
SUTRA HiFi
- Purpose: Voice-to-voice interaction with natural, low-latency responses.
- Features: Dual Diffusion-Transformer architecture, supports 30+ languages, ultra-low latency for real-time voice applications.
- Use Cases: Virtual assistants, customer service IVR, multilingual voice bots.
📦 Step 1: Install Dependencies
# SUTRA models are OpenAI API compatible
!pip install -qU openai🔐 Step 2: Example with SUTRA-V2
from openai import OpenAI
client = OpenAI(
api_key="YOUR_SUTRA_API_KEY",
base_url="https://api.two.ai/v2"
)
response = client.chat.completions.create(
model="sutra-v2",
messages=[{"role": "user", "content": "Explain AI in Tamil."}],
max_tokens=1024,
temperature=0.7
)
print(response.choices[0].message.content)🛠 Troubleshooting
- Model Selection: Use correct model names (
sutra-v2,sutra-r0, etc.). SUTRA-V1 was deprecated on March 22, 2025. - Invalid API Key: Verify your key at https://developer.two.ai.
- Rate Limits: Reduce request frequency or contact TWO AI at https://www.two.ai/support.