🔑
Bring your own keys
Keep direct provider billing and avoid model-call markup while RouteIQ handles the gateway layer.
RouteIQ · AI Gateway
Bring your own LLM keys. We run the gateway.
RouteIQ is AI gateway infrastructure for developers, startups, and enterprises building AI-powered applications. Route across providers, track latency and usage, support MCP traffic, and scale without rebuilding your stack.
AI gateway
One entry point across providers, policies, and agent workflows.
Keep direct vendor billing. Centralize routing, fallback, and operational visibility.
OpenAIHealthy · 162ms
AnthropicFallback ready
MCP toolsObserved
Key Benefits
Ship AI faster without giving up control of your stack.
🛰️
Route across models
Stay model-agnostic with provider routing, fallback behavior, and policies outside your app code.
📡
See every request
Track latency, errors, usage, and agent traffic from one operational surface instead of many logs.
Features
Everything teams need from a serious AI gateway.
Keys and access
Keep your provider accounts. Centralize access above them.
RouteIQ sits between your apps and model providers without taking over billing or forcing a single-vendor AI strategy.
- Use your own API keys across supported LLM providers.
- Keep direct vendor billing with no markup on model calls.
- Manage access by team, app, or environment.
Routing and reliability
Route every request to the right model.
Handle multi-LLM routing, fallback, and reliability decisions once instead of scattering them across application code.
- Route by policy, latency, cost, model family, or use case.
- Fail over cleanly when a model or provider becomes unreliable.
- Keep application code simpler while routing strategy evolves.
Routing policy
Default to fastest model. Fail over on p95 spikes.
Traffic routed this hour14.2k requests
Fallback events0.8%
Provider diversity3 active vendors
Observability and agents
Observe chat, tool, and MCP traffic from the same layer.
Run chat, tool, and MCP workflows through the same gateway so teams do not need separate instrumentation for core AI traffic and agent calls.
- Track latency, errors, throughput, usage, and provider performance in one place.
- Support MCP servers and tool-driven agent workflows.
- Scale the same AI gateway from product launch to enterprise rollout.
p95 latency178ms
Error rate0.24%
Observed MCP calls2.1k today
Use Cases
Made for teams turning AI into product infrastructure.
Developers
Stop rebuilding gateway plumbing
Ship faster with LLM gateway abstraction, routing, and observability already handled.
Startups
Move faster without lock-in
Experiment across model vendors while keeping one AI gateway layer as your traffic grows.
Enterprises
Standardize AI traffic
Create one AI gateway policy and telemetry layer across multiple apps, environments, and teams.
How It Works
Four steps from provider sprawl to controlled AI traffic.
Connect providers
Add your own LLM keys and environments.
Define policies
Set routing, failover, and access behavior outside app code.
Send traffic
Run chat, tool, and MCP requests through one gateway endpoint.
Observe and tune
Track usage, latency, and failure patterns as volume grows.
Pricing teaser
Pay providers for model usage. Pay RouteIQ for infrastructure.
RouteIQ pricing covers the gateway layer: multi-LLM routing, LLM observability, policy, and scalability. Your model usage stays with the providers you already pay.
Provider billingDirect
Markup on callsNone
What RouteIQ coversGateway infrastructure
Trust and Differentiation
Why teams pick RouteIQ over DIY gateway work.
💸
No markup on calls
Keep direct provider economics while Teptro handles the layer above them.
🔀
Multi-LLM routing
Evolve providers without rewriting product logic every quarter.
👁️
Observability by default
See usage, latency, and failure patterns without a side project.
🧰
MCP and agent support
Treat agent workflows as first-class traffic instead of special cases.
🪢
Less vendor lock-in
Keep freedom to shift providers, policies, or model mix as the market changes.
📈
Built to scale
Use one gateway from early launches through enterprise rollout.
Pricing
Configure RouteIQ for your AI traffic.
Pick request volume, provider count, and governance features. Pricing adjusts live based on your AI gateway needs.
Build on your own LLM stack. Let RouteIQ run the gateway.
Start with the provider accounts you already have and add production routing, LLM observability, MCP support, and scale from day one.
FAQ
RouteIQ questions.
RouteIQ is Teptro's AI gateway infrastructure for teams building AI-powered products with their own LLM provider keys.
Yes. RouteIQ is designed around bring-your-own-keys so you keep direct vendor billing, pricing, and commercial control.
Yes. Multi-LLM routing is a core part of the platform, including routing, fallback, and provider flexibility as your model strategy evolves.
Yes. RouteIQ is designed to support agent traffic, tool calls, and MCP workflows so the same gateway layer can handle advanced AI operations.
You use RouteIQ for AI gateway infrastructure: routing, observability, policy, and scalability. Model usage stays with your chosen providers.
It is built for developers, startups, and enterprises building AI-powered applications that need multi-LLM routing, LLM observability, and a gateway layer that can scale.