Best LLM routers
LLM routers choose the model or tier for each request. Use this page when cost, quality, latency, or reliability depends on routing policy rather than one static model pick.
Filtered router list
Heureka Labs UG
Passthrough + fee
Commercial LLM router that analyzes incoming requests and routes to the optimal model for cost/quality/latency via a drop-in OpenAI-compatible API, with a privacy-preserving embedding mode that avoids sending prompt content.
Amazon Web Services
Passthrough
AWS Bedrock's native intelligent prompt router that routes prompts between Anthropic Claude model tiers (Haiku/Sonnet) based on predicted task complexity, with no extra per-routing charge.
Microsoft
Passthrough
Microsoft Azure AI Foundry's native model router that uses a trained ML model to route each prompt in real time to the optimal Azure-hosted model, with Balanced/Cost/Quality mode selection and automatic failover.
Martian, Inc.
Passthrough + fee
AI-powered LLM router that analyzes each prompt in real-time to select the optimal model, targeting 20–97% cost reduction while maintaining quality; San Francisco startup reportedly nearing $1.3B valuation.
Neutrino AI
Passthrough + fee
Commercial LLM router that dynamically routes each query to the best-suited model with load balancing and fallback handling, charging 3% of underlying AI spend.
Not Diamond
Enterprise quote
Predictive model router that determines the best LLM for each query; claims up to 25% accuracy gains and 10x cost reduction; powers OpenRouter's auto mode and is positioned specifically for coding agents.
Passthrough
OpenAI's native auto-routing mode (GPT-5 Auto) that dynamically routes each API request between GPT-5 and GPT-5 Instant based on prompt complexity, with no extra charge beyond model token costs.
LMSYS (lm-sys)
Free OSS
Open-source LLM routing framework from LMSYS that routes simpler queries to a cheaper weak model and harder ones to a stronger frontier model, achieving 35–85% cost reduction on benchmarks.
Unify AI
Subscription
Benchmark-driven LLM router using a neural scorer and live runtime benchmarks refreshed every 10 minutes to route each request to the optimal endpoint across 100+ providers.
Red Hat / vLLM Project
Free OSS
Open-source Mixture-of-Models router that semantically classifies each request and routes it to the best backend (local, private, or frontier) by cost, latency, privacy, or safety, deployed as an Envoy External Processor.
Related decision paths
These links keep the SEO view tied to the single router store, target-provider graph, and model decision pages.
Compare every verified gateway, router, and hybrid row in the shared store.
Use the directory controls to add openness, hosting, API, and routing-scope filters.
Inspect the provider that one or more listed routers can target.
Inspect the provider that one or more listed routers can target.
Inspect the provider that one or more listed routers can target.
Inspect the provider that one or more listed routers can target.
Use model-level pricing when routing is not needed.
Start with task quality before adding a routing layer.
Check task fit for agentic workflows before routing.
Machine-readable source
This page is generated from data/seed/router.json. Agents can consume the same catalog through /api/routers.