The LLM gateway landscape
Every LLM gateway.
Every model they serve.
One place to compare.
4,294 models across 9 live LLM gateways — who carries what, with what context window. Crawled from each gateway's own catalog, not hand-curated.
Availability data only in this snapshot.
This build was generated from FreeRouter's model-map snapshot, which records
which gateway serves which model but carries no pricing, context windows, or
descriptions. Those columns fill in once the catalog refresh runs against the
gateway APIs — see about the data.
4,294
Models catalogued
9
Live gateways
8 more tracked
154
Model vendors
437
On 2+ gateways
Routable alternatives
—
Largest context
The gateways
Nine live gateways, and the ones we are watching
A gateway is "live" here when we hold a catalog crawled from its own API. The others get a profile page and no invented data.
Compare all gatewaysMost portable
Models with the most gateway options
The more gateways carry a model, the less any one of them can hold your application hostage.
Browse every model| Model | Vendor | Type | Gateways | Context |
|---|---|---|---|---|
claude-fable-5
anthropic/claude-fable-5
|
Anthropic | Unclassified | 5 | — |
claude-fable-5.1
anthropic/claude-fable-5.1
|
Anthropic | Unclassified | 5 | — |
claude-opus-5
anthropic/claude-opus-5
|
Anthropic | Unclassified | 5 | — |
claude-sonnet-5
anthropic/claude-sonnet-5
|
Anthropic | Unclassified | 5 | — |
deepseek-v4-flash-0731
deepseek/deepseek-v4-flash-0731
|
DeepSeek | Unclassified | 5 | — |
deepseek-v4-pro-0813
deepseek/deepseek-v4-pro-0813
|
DeepSeek | Unclassified | 5 | — |
gpt-4.1
openai/gpt-4.1
|
OpenAI | Unclassified | 5 | — |
gpt-4.1-mini
openai/gpt-4.1-mini
|
OpenAI | Unclassified | 5 | — |
gpt-4.1-nano
openai/gpt-4.1-nano
|
OpenAI | Unclassified | 5 | — |
gpt-4o
openai/gpt-4o
|
OpenAI | Unclassified | 5 | — |
gpt-4o-mini
openai/gpt-4o-mini
|
OpenAI | Unclassified | 5 | — |
gpt-5
openai/gpt-5
|
OpenAI | Unclassified | 5 | — |
gpt-5-mini
openai/gpt-5-mini
|
OpenAI | Unclassified | 5 | — |
gpt-5-nano
openai/gpt-5-nano
|
OpenAI | Unclassified | 5 | — |
gpt-5.2
openai/gpt-5.2
|
OpenAI | Unclassified | 5 | — |
Browse
By model type
Gateway catalogs are not only chat models — they carry image, video, audio, embedding, and rerank models too.
Browse
By vendor
Who actually makes the models, regardless of which gateway resells them.
All vendorsQuestions
About gateways, models, and this data
An LLM gateway is a service that sits between your application and the model providers, exposing many models behind one API — usually OpenAI-compatible. Instead of integrating OpenAI, Anthropic, and Google separately, you call the gateway and name a model. Gateways differ in which models they carry, what they charge, and what they add on top: caching, failover, spend controls, or routing policies.
Each gateway sets its own markup over the underlying provider, and some let you bring your own provider key and pay the provider directly. Add differing volume agreements and the same model id can carry meaningfully different prices depending on where you send the request.
Every catalog here is crawled from the gateway's own /models endpoint by FreeRouter, normalized onto OpenRouter-style canonical model ids, and served through FreeRouter's public catalog API. Nothing is hand-entered, and each gateway page shows when its catalog was last checked.
No. They are what each gateway published in its own catalog at the time of the last crawl, normalized to US dollars per million tokens. Treat them as a directional comparison and confirm against the gateway before committing to anything. A blank cell means the gateway publishes no price for that model, not that it is free.
No — and that is the interesting part. 437 of the models listed here are served by more than one gateway. A meta-router such as FreeRouter puts one API key in front of several gateways at once, so you can route by price or latency and fail over automatically rather than picking a single vendor up front.