See how ChatGPT, Perplexity and Google AI Overviews describe you today
Free AI Visibility AuditSwitchyard is an open-source model router from NVIDIA that sits in front of multiple models and picks which one handles each request, with cost as the thing it optimises for.
01 / decision
Each figure sits next to the middle of the field, so it can be read as dear or cheap, wide or narrow, rather than floating on its own. Compared against the 232 models we track, not against an absolute standard.
Capability
Not measured
No published benchmark scores yet
Input / 1M tokens
Not listed
field median $0.41
Output / 1M tokens
Not listed
field median $2.00
Context
1,000K
field median 500K
02 / overview
Four questions, answered separately, because somebody arrives at one of them rather than at the top of the page.
The job it was built for, and the job it is not for.
Switchyard is routing infrastructure, not a model that generates anything of its own. Its job is to take an incoming text request and decide which underlying model should answer it, using OpenRouter market data by default to select from the most popular options. It is not something you prompt for reasoning, writing or analysis, the quality of what comes back is the quality of whichever model it routed to.
Where it sits in its line, and when a sibling is the better pick.
Switchyard first appeared in September 2026 from NVIDIA, as an open-source release rather than a hosted product tier. It is the right pick when you are already running traffic across several models and want the choice made per request instead of hard-coded. It is the wrong pick if you are committed to one model, or if you need predictable, identical behaviour on every call, since routing by definition varies what answers you.
The API, the apps it powers, and what its limits let you do.
Switchyard is open source, so you run it yourself in front of your model calls rather than hitting someone else's endpoint. It handles text, and its context allowance is large enough that routing decisions are not the thing constraining how much you can send through. Out of the box it leans on OpenRouter market data to decide where a request should land, which you can change.
What changes because this exists, or why it does not.
Routing between models used to be something teams wrote and maintained themselves, usually as a pile of conditionals that went stale as prices and model lineups moved. Switchyard makes that layer a standard component you can adopt rather than build, with the market data feed doing the part that otherwise needs manual upkeep. For most teams this is plumbing, useful, not dramatic.
03 / evidence
Split by category, so a strong number never hides a thin evidence base. Verified means we read it on the benchmark's own published results; a provider's claim about its own model is shown and labelled rather than dropped.
No published benchmark scores for this model yet.
We publish a score only where we can link the result to where it was published. Until a benchmark result for this model exists in a source we read, this section stays empty rather than being filled with a provider's marketing figure.
How we decide what counts as evidence04 / ledger
Every published row, grouped by category, each compared with the best published score on the same benchmark. 'Is 64% good' is a question nobody can answer; '26 points behind the leader' is one anybody can.
Nothing in the ledger yet.
Each row here carries a score, the benchmark it came from, what the leading model scored on the same test, and a link to the published result. Rows appear as results are published and read.
How we decide what counts as evidence05 / capability
Where this model is strong, and against how many peers. Ranks are against models with evidence in that category, not against everything we track: ranking against models nobody tested would rank who published, not who is better.
No category scores to shape yet.
A category score is the weighted mean of the benchmarks published for it. With no published rows there is nothing to average, and an empty chart drawn at zero would say something false.
How we decide what counts as evidence06 / cost
List API rates as last read from the provider, with the source on every row, plus every change we have recorded since we started tracking it.
No price recorded for this model.
Either the provider publishes no per-token rate for it, or we have not read one yet. We would rather show nothing than an estimate you cannot check against the provider's own page.
How we decide what counts as evidence07 / specs
As listed by the provider's own catalogue and re-read every few hours. Anything absent is absent there too.
| Context window | 1,000,000 tokens |
|---|---|
| Maximum output | Not listed |
| Modalities | text |
| Released | 21/09/2026 |
| Status | Current |
| Catalogue identifier | nvidia/switchyard |
08 / lineage
What this model replaced, what replaced it, and what else its provider has in the field.
Also from NVIDIA
09 / line
Every model in this family in release order, so a page from eight months ago says in one glance that two newer ones exist.
Nothing else in this line yet.
A line is worked out from the naming and the release dates across every model page we hold. It fills in as the provider ships successors, or as we pick up the models that came before this one.
How we decide what counts as evidence10 / notes
What this model changes for a brand trying to be cited in AI answers, and every change we have logged since it launched.
What it changes for you
Routers like Switchyard mean the model answering a question about your category may not be the model the product nominally runs on, and may change between requests. Visibility testing that checks one model is a weaker signal than it used to be, you want coverage across the set a router would plausibly choose from. The practical implication is the same as always, be the source that any competent model cites, rather than tuning for one.
Buyers do not meet Switchyard directly, there is no consumer app and no chat surface carrying its name. It shows up inside products other people build, deciding quietly which model answered the question a user just asked. If you are evaluating it, you meet it as a repository and a config file, not as an interface.
Name: Switchyard Provider: NVIDIA Listed: 2026-09-21 Context window: 1000000 tokens Switchyard is an open-source model router that switches between multiple models to optimize the cost of requests. By default it will use OpenRouter market data to select the most popular...
11 / questions
The things people ask about this model, answered from what is on this page rather than from anywhere else.
Models change what gets cited. We measure whether AI answers name your brand or your competitors across every assistant, and show you what to change.