One OpenAI-Compatible Router.
Every Frontier & Open-Source Model.
A single intelligent endpoint providing access to all frontier reasoning engines and top-tier open-source models. Zero rate-limit hang-ups, fully coding agent compatible (Cursor, Continue, Cline, Roo Code), with transparent billing — a monthly allowance from ₹499 or pure pay-for-what-you-consume credit from ₹99 (free 2K trial included).
One Intelligent Router for All Models
Send model as "auto". The router inspects request tokens, intent complexity, and real-time provider health to dispatch the optimal model with automatic 25% burn fallback.
Deep Reasoning & Math
Dispatches multi-step mathematical problems, algorithmic synthesis, and complex planning to frontier reasoning engines.
deepseek-v4-pro Active
qwen3.8-max Active
kimi-k3 262kqwen3.7-max Activeqwq-plus Reasoning
Agentic Programming
Specialized code models tuned for repository refactoring, bug isolation, type check validation, and CLI coding agents.
qwen3-coder-plus Activekimi-k2.7-code Activeqwen-coder-plus Active480b-moe-coder Active
Ultra-Low Latency
Sub-second response stream for conversational bots, high-throughput classification, structured JSON extractions, and summaries.
qwen3.8-flash <1sdeepseek-v4-flash Activeqwen-turbo Activeqwen3.6-27b Lowdeepseek-v3.2 Active
Multimodal Vision & OCR
Processes UI screenshots, architectural diagrams, medical charts, and dense document OCR automatically.
qwen3-vl-plus 128kkimi-k3 Visionqwen-vl-max Activeqwen-vl-ocr OCRqwen3.5-omni OmniTop 5 models in use right now
A monthly allowance. Overage at cost.
Pick a monthly plan for a large token allowance at a discount, and anything you use beyond it draws from wallet credit at the flat pay-as-you-go rate — so you are never cut off mid-build, and never billed for headroom you did not touch. Your first 2,000 tokens are free, no card required.
Pay For What You Consume — Zero Minimum Commitments
1st Recharge ₹99 • Then Pay Anything (₹50, ₹90+)Flat ₹99 per 1M tokens across every model (10,000 tokens per ₹1). Pay just ₹99 once to activate your API keys with 1,000,000 tokens. After that initial deposit, there are zero minimum charges — deposit ₹50, ₹90, or whatever you want! Tokens never expire.
Coding Agent & OpenAI SDK Compatible
Integrates out of the box with Cursor IDE, Continue.dev, Cline, Roo Code, LangChain, and the official OpenAI Python/Node.js SDKs. Just configure base URL.
from openai import OpenAI # Point client to your LLM Pool Proxy instance client = OpenAI( base_url="http://127.0.0.1:3001/v1", api_key="sk-live-your_custom_api_key" ) # The single router automatically shifts models based on quota & complexity response = client.chat.completions.create( model="auto", # Or target "auto-code", "auto-complex", "deepseek-v4-pro" messages=[ {"role": "user", "content": "Refactor this auth middleware to use ed25519 signatures"} ] ) print(response.choices[0].message.content)
// Cursor IDE / Continue.dev / Cline Custom Model Configuration
{
"models": [
{
"title": "LLM Pool Auto Router",
"provider": "openai",
"model": "auto-code",
"apiBase": "http://127.0.0.1:3001/v1",
"apiKey": "sk-live-your_custom_api_key"
}
]
}