Welcome to SUPREME AI Developer Platform
The SUPREME Setu Gateway provides high-throughput, low-latency access to frontier artificial intelligence models through an industry-standard OpenAI-compatible API interface and the open Model Context Protocol (MCP).
Whether you are building conversational agents, educational tutors, reasoning engines, or multimodal applications, SUPREME Setu serves the sovereign SUPREME Quantum model family through resilient Anycast South Asian edge nodes.
Core Capabilities
- Unified OpenAI Interface: Drop Setu directly into existing OpenAI SDKs (Python, TypeScript, Go), LangChain, or Vercel AI SDK by setting
baseURL. - Native Model Context Protocol (MCP): Connect enterprise databases (Postgres, Firestore), ERPs (SAP, Tally), CRMs (Salesforce), and local filesystems directly to models.
- Intelligent Fallback Chains: Automatic zero-downtime failover between foundation clusters when upstream latency spikes or outages occur.
- Devanagari BPE Efficiency: 4.2x token compression for Nepali and Hindi text compared to standard Western tokenizers, reducing cost by up to 76%.
- Transparent Credit Economics: Every request is metered transparently in Nepali Rupees (NRs) and USD with zero hidden minimums.
Base URL & Gateway Address
All API requests are routed through the secure Setu gateway:
bash
https://supremeai.services/api/setu/v1For OpenAI SDK integrations, simply configure baseURL:
typescript
import OpenAI from 'openai';
const client = new OpenAI({
apiKey: process.env.SUPREME_API_KEY, // sup_live_...
baseURL: 'https://supremeai.services/api/setu/v1',
});Available Models
| Model ID | Target Workload | Context Window | Best For |
|---|---|---|---|
quantum-lite | Sub-150ms Reflexive Routing | 128k tokens | High-throughput classification, intent extraction, and automated workflows |
quantum-flash | Multimodal Audio & Vision | 256k tokens | Real-time conversational voice loops, optical document parsing, and UI generation |
quantum-reasoner | Dynamic Test-Time Compute (TTC) | 128k tokens | Complex mathematics, physics derivations, algorithmic proofs, and Socratic tutoring |
quantum-pro | Enterprise Deep Context | 1,000,000+ tokens | Multi-repository refactoring, institutional legal compliance, and vast document archives |
Next Steps
- Follow the Quickstart Guide to send your first completion request in under two minutes.
- Learn about Authentication & API Keys to create scoped keys (
sup_live_,sup_test_). - Review Rate Limits & Quotas for RPM and concurrency ceilings.
- Check the API Error Codes & Fallback Matrix for deterministic error handling.
- Explore the Chat Completions API Reference for full request schema and parameters.