Intelligent routing
Rule-based and optional LLM classifiers send easy prompts to small models and complex ones to premium tiers. Failover keeps chat alive when a vendor is down.
SmartRoutIQMulti-LLM gateway
SmartRoutIQ
One intelligent gateway that routes every prompt to the right model — faster when it can, stronger when it must.
OpenAI
Claude
Gemini
Llama
Mistral
DeepSeek
SmartRoutIQ
Intelligent Multi-LLM Gateway with Smart Routing
OpenAI
Claude
Gemini
Llama
Mistral
DeepSeek
SmartRoutIQ
Intelligent Multi-LLM Gateway with Smart Routing
How SmartRoutIQ works
Bring your provider keys. We classify the task, pick a model, call it through LiteLLM, and keep cost, latency, and safety in view.
- 01
Connect your providers
Add OpenAI, Anthropic, Groq, OpenRouter, and more with your own API keys. One gateway, many models.
- 02
SmartRoutIQ routes the ask
Each prompt is classified by complexity. Easy work goes to fast cheap models; hard work goes to stronger ones — with fallback if a provider fails.
- 03
Chat, RAG, and govern
Use Chat and RAG Chat in the app, track cost and latency, and keep guardrails on secrets and budgets.
Product walkthrough
A short animated demo of connectors, routing, and Chat — built into the site, no YouTube required.
SmartRoutIQ+ OpenAI
+ Groq
+ Anthropic
+ OpenRouter
Connecting…
SmartRoutIQ · Demo
Connect providers
Add OpenAI, Groq, Anthropic, OpenRouter — your keys, one gateway.
What you get
Your keys, your spend
Connectors store provider credentials per account. Platform plans cover the gateway; model usage is billed by your providers.
Chat + RAG
Built-in Chat for day-to-day routing. Pro unlocks RAG Chat against Weaviate / hosted indexes with API embeddings.
Governance
Budget caps, secret scanning, request logs, and plan limits so teams stay inside policy without bolting on another tool.