Two products keep hiding under one word If you are shopping for a way to stop sending every LLM request to the most expensive model, two names come up fast: vLLM Semantic Router and Nadir. They make the same promise, spend less without losing quality, and they both call themselves routers. That shared word hides the actual decision. One is an open source gateway you install and operate. The other is a managed decision layer you point your existing code at. Choosing between them is not a benchmark question. It is a build-versus-buy question, and the honest answer depends on who is going to own the thing at 2am. This is a straight comparison: what each one is, what each one costs to run, where each one wins, and how to tell which side of the line your team is on. We are not going to pretend one of these is bad. They are both good. They are good for different teams. What vLLM Semantic Router is vLLM Semantic Router (VSR) is an open source project from Red Hat's CTO office and the vLLM community, Apache-2.0 licensed, that launched in September 2025 and crossed 2,000 GitHub stars within two months. Source: Red Hat, "Bringing intelligent, efficient routing to open source AI with vLLM Semantic Router". It runs as an Envoy External Processor: a gRPC filter, written in Go for the network path and Rust for the ML path, that inspects every prompt before Envoy forwards it and decides which backend model should answer. Source: Red Hat Emerging Technologies, "Intelligent inference request routing for large language models," November 11, 2025. By its Athena release it had grown into eight neural classifiers on a 307-million-parameter mmBERT backbone (intent, jailbreak, PII, hallucination, domain, complexity, safety, and fact-check), a Boolean signal-decision engine, an HNSW semantic cache, and eleven model-selection algorithms. Source: Red Hat Developer, "Getting started with the vLLM Semantic Router project's Athena release," March 25, 2026. Its own published benchmark is strong and worth respecting: routing reasoning mode on or off per query on MMLU-Pro with Qwen3-30B cut tokens 48.5%, cut latency 47.1%, and raised accuracy 10.2%. Source: Red Hat Developer, "vLLM Semantic Router: Improving efficiency in AI reasoning," September 11, 2025. Open sourcing this category is a genuinely good outcome for the industry. What it is not is free to operate. The documented production footprint is Kubernetes, Envoy, vLLM, Hugging Face model infrastructure, and Prometheus with Grafana, shipped through Helm charts and two custom resources (IntelligentPool and IntelligentRoute), with roughly 25GB of classifier models kept warm before it routes a single request. Source: Red Hat Developer, "Getting started with the vLLM Semantic Router project's Athena release," March 25, 2026. Open source describes the license, not the running cost. Someone on your team owns a Kubernetes-native gateway, an eight-classifier ML pipeline, and a monitoring stack, across major releases that landed roughly every two months this year. We wrote a full breakdown of that operational cost here. What Nadir is Nadir is a hosted decision layer that speaks the OpenAI API. You change one base URL, set the model to auto, and Nadir reads each request, picks the cheapest model that can answer it well, and returns the completion in the same call. There is no cluster to stand up, no classifier to host, and no client rewrite, because the request and response shapes are the ones your code already uses. from openai import OpenAI The only change: the base URL, and model="auto" client = OpenAI(base_url="https://api.getnadir.com/v1", api_key="YOUR_NADIR_KEY") resp = client.chat.completions.create( model="auto", # Nadir picks the model per request messages=[{"role": "user", "content": "Summarize this contract clause..."}], ) Under that endpoint is a learned routing brain plus a verifier. The router classifies difficulty, routes to a model tier, and a calibrated verifier scores the cheap model's answer before it ships, escalating only when the score is too low. On 11,420 RouterBench held-out triples this cut cost 60% against always using the top model while preserving about 98% of its quality. The point of the verifier is that the router is graded on the answer it actually produced, not on a guess it made before seeing one. Routing without that check is dead reckoning. Around the routing decision, Nadir ships the parts VSR leaves to you: a cost and savings dashboard, per-request logs, fallback chains, spend controls, and billing. For teams with a compliance boundary, Nadir also self-hosts and runs on-prem, so keeping classification inside your own network does not force you onto a build-it-yourself gateway. The real question is build versus buy Both projects route. The difference is what you have to own to get there. VSR hands you a powerful, controllable substrate and asks you to operate it. Nadir hands you the outcome and operates the substrate for you. | | vLLM Semantic Router | Nadir | |---|---|---| | What it is | Open source Envoy filter you self-host | Managed decision layer behind an OpenAI-compatible endpoint | | License | Apache-2.0, free | Free with BYOK; usage-based on the hosted plan | | Infrastructure you own | Kubernetes, Envoy, classifier hosting, Prometheus/Grafana | None; change one base URL | | Time to first routed request | Cluster, Helm, CRDs, then warm ~25GB of models | A base URL change and model=auto | | Routing brain | You train, host, and upgrade the classifiers | Maintained and versioned for you | | Published quality result | Strong reasoning-mode benchmark (48.5% fewer tokens); a routing-quality target on your own traffic is yours to establish | 60% lower cost at about 98% of quality preserved on 11,420 RouterBench held-out triples | | Verification before shipping a route | Not the project's stated focus | Verifier-gated: the cheap answer is scored before it ships | | Observability and billing | You assemble it | Cost dashboard, per-request logs, spend controls, billing included | | Compliance and on-prem | In-network by design | Hosted, or self-host and on-prem when classification must stay inside your boundary | | Best fit | Teams already operating vLLM plus Kubernetes at scale | Teams whose calls already hit hosted APIs and want routing without a platform team | Both columns are legitimate answers to the same question. They are answers for different teams. If you are already running a vLLM fleet behind Kubernetes and Envoy, the marginal cost of adding VSR is low and full control over the classifier is a feature. If your traffic already goes to hosted frontier APIs and nobody wants to own a new Kubernetes gateway, the routing is a base URL change away. You may not have to choose the layer The table reads as an either/or, but the line is softer than that. Nadir is also available as a provider you can drop into an existing gateway such as LiteLLM, which means a team that already standardized on a gateway can keep it and let Nadir be the routing brain behind it. If you have already invested in Envoy and Kubernetes, you are not throwing that away to get a managed routing decision. The build-versus-buy line sits at the routing brain and the operational surface, not at your whole gateway. Keep the carrier you like, and let the decision layer be the part with a measured guarantee. The part that quietly decides quality Every router makes the same bet: this request is easy enough for a cheaper model. The gap between a good router and a bad one is what happens when that bet is wrong. A predictive router that classifies difficulty and routes, with no check on the result, ships whatever the cheap model produced, including the times it was over its head. That is where quality loss hides, not in the average case but in the tail. Nadir closes that tail with a verifier cascade. The cheap model answers, a calibrated scorer grades that answer, and only a low score escalates to a stronger model. The router is measured on the answer it returned, not the guess it made before generating one. That is the mechanism behind the 60% cost cut at about 98% of quality preserved: the savings come from routing down aggressively, and the quality comes from catching the misroutes before they reach the user. VSR gives you real signals to build a policy like this, its complexity and hallucination classifiers among them, but assembling a verified cascade and calibrating the escalation threshold on your own traffic is a project you run. With Nadir it is the default behavior of the endpoint. If you want the longer argument for why the check matters more than the classifier, we made it here. Which one fits your team Choose vLLM Semantic Router if you already operate vLLM, Kubernetes, and Envoy at scale, you need classification to stay fully in-house for compliance, and you have platform engineers who can own a gateway that shipped three major versions in five months. Full control is the payoff, and for you it is worth the operating cost. Choose Nadir if your traffic already hits hosted APIs like Claude, GPT, and Gemini, you want routing live this afternoon without a hiring plan, and you want the savings, the verification, the logs, and the billing as one product rather than five systems you wire together. Consider both if you want a gateway you control and a routing brain with a measured guarantee. Run Nadir behind your gateway and you get the operational control of one and the verified, published economics of the other. Bottom line vLLM Semantic Router is a strong open source project, and if your team already lives in vLLM, Kubernetes, and Envoy, adopting it is a reasonable move with full control as the reward. What it asks in return is that you own the gateway, the classifiers, the monitoring, and the upgrade treadmill. Nadir is the same routing idea sold as an outcome instead of a construction project: one base URL change, a learned router with a verifier, a published 60% cost reduction at about 98% of quality preserved, and the dashboard and billing already running. If you would rather ship routing today than staff it, that is the trade Nadir is built to make. And if you want both, a gateway you control and a routing brain with a guarantee, Nadir runs behind your gateway too. Point your base URL at Nadir, set model=auto, and watch the savings land on the dashboard from the first request. Start free, or read the operational-cost breakdown of self-hosting vLLM Semantic Router if you are leaning toward running it yourself. If you are still deciding whether routing belongs in your stack at all, the LLM routing guide starts one level up. The vLLM Semantic Router figures in this piece are drawn from the project's own published benchmarks and documentation, not from Nadir's traces. Nadir's 60% cost and 98% quality figures are measured on 11,420 RouterBench held-out triples. Sources: Red Hat Developer, "vLLM Semantic Router: Improving efficiency in AI reasoning," September 11, 2025. Red Hat, "Bringing intelligent, efficient routing to open source AI with vLLM Semantic Router". Red Hat Emerging Technologies, "Intelligent inference request routing for large language models," November 11, 2025. Red Hat Developer, "Getting started with the vLLM Semantic Router project's Athena release," March 25, 2026. GitHub, vllm-project/semantic-router. vLLM Semantic Router homepage.