Retrieve a model
One model, by its id or alias, as the model list shows it.
https://inference.trymorse.ai/v1/models/{id}Retrieves one model, as OpenAI's model retrieve does, with the entry GET /v1/models lists for it to this caller. A model id, such as z-ai/glm-5.3, which holds a slash, answers its entry with its tiers; an alias, such as glm-5.3, answers its entry at its tier. The call is free, needs no balance, and shares the account's 60 model lists a minute.
curl https://inference.trymorse.ai/v1/models/moonshotai/kimi-k3 \
-H "Authorization: Bearer $MORSE_API_KEY"{
"id": "moonshotai/kimi-k3",
"object": "model",
"created": 0,
"owned_by": "morse",
"type": "chat",
"label": "Kimi K3",
"input_modalities": [
"text",
"image"
],
"reasoning_efforts": [
"none",
"low",
"high",
"max"
],
"tiers": [
{
"tier": "private",
"retention": "zdr",
"quantizations": [
"mxfp4",
"fp8"
],
"context_tokens": 1048576,
"max_output_tokens": 1048576,
"compact_at_tokens": 838860,
"default_reasoning_effort": "high",
"plan": "pro",
"price": {
"prompt_micros_per_mtok": 3000000,
"completion_micros_per_mtok": 15000000,
"min_hold_micros": 15360
}
},
{
"tier": "e2ee",
"retention": "vendor",
"attesters": [
{
"name": "phala",
"sealed_model": "moonshotai/kimi-k3"
}
],
"quantizations": [
"unknown"
],
"context_tokens": 1048576,
"max_output_tokens": 1048576,
"compact_at_tokens": 838860,
"plan": "pro",
"price": {
"prompt_micros_per_mtok": 3000000,
"completion_micros_per_mtok": 15000000
}
}
]
}Response
The model, as GET /v1/models lists it to this caller: a model id's entry with its tiers, or an alias's entry at its tier.
idstringRequiredThe model's id.
object"model"RequiredcreatedintegerRequiredAlways 0.
owned_bystringRequiredtypestringRequiredThe endpoint that serves the model, as for a model in the aliases shape.
One of
chat,image,video,search,fetch,research,transcription, orspeech.labelstringinput_modalitiesarray of stringAbsent when the model reads text only.
reasoning_effortsarray of stringOn every chat model: the efforts it takes, lowest first. Empty when it takes none.
One of
none,minimal,low,medium,high,xhigh, ormax.always_reasonstrueTrue on a chat model that reasons on every request and takes no effort, whose reasoning_efforts is empty. Absent on every other model.
toolsfalseFalse on a chat model served without tools. An API key's request that lets it call a tool (tools, or a tool_choice that calls one, with a tool_choice other than none) is refused with 400 unsupported_parameter naming tools; any other request reaches the host with no tools, tool_choice, or parallel_tool_calls. Absent on every model that takes tools.
charactersfalseFalse on a chat model kept out of character chats. A chat that plays a character on it, one that sends X-Character or a character's prompt, is refused with 400 model_unsupported_for_characters before Morse calls the model or holds any credit. Absent on every model that takes character chats.
tiersarray of ModelTierRequiredThe tiers the model is sold at, from the least private to the most. Each tier appears once, whichever providers serve it.
Show 14 fields
tierstringRequiredThe privacy tier. anon reaches the vendor without your identity, private runs on a provider with zero data retention, tee runs on attested hardware, and e2ee also encrypts the messages to that hardware.
One of
anon,private,tee, ore2ee.retentionstringOne of
vendororzdr.attestersarray of objectOn a tee or e2ee row: the vendors whose attestation proves the tier. On an e2ee row each also gives the model a sealed body carries.
quantizationsarray of stringOn a chat row: the precisions the tier's providers publish for the weights, in OpenRouter's names and order: integers, then floats, each fewest bits first, and unknown last. unknown is a provider that publishes none. The list counts every provider a chat at the tier may run on, backups and e2ee fallbacks included, and names none of them. provider.quantizations chooses among these.
One of
int4,int8,fp4,mxfp4,nvfp4,fp6,fp8,mxfp8,fp16,bf16,fp32, orunknown.context_tokensintegermax_output_tokensintegerThe most output tokens a request served at this tier may produce. A request asking more gets this many below e2ee and is refused with e2ee_output_limit_refused at e2ee.
compact_at_tokensintegerinput_modalitiesarray of stringWhat a chat at this tier reads, present only when it differs from the model's input_modalities: ["text"] on a text-only tier of a model that reads images. A client offers the row's value when it has one and the model's otherwise.
reasoning_effortsarray of stringThe efforts a chat at this tier takes, lowest first, present only when they differ from the model's reasoning_efforts. A client offers the row's value when it has one and the model's otherwise.
One of
none,minimal,low,medium,high,xhigh, ormax.default_reasoning_effortstringThe effort a chat gets when it sets none. Absent when the vendor chooses.
planstringThe lowest plan the Morse app lists the model under.
requires_planstringThe lowest plan that may call the model in the Morse app. It does not apply to API keys.
includedobjectWhat plans include in the Morse app. It does not apply to API keys.
priceobjectRequiredWhat you pay at this tier, in micro-dollars. 10,000 micros are one credit, which is $0.01. An API key pays the vendor's list price, and a Morse app login pays it at the model's markup. A chat model lists its prompt and completion price per million tokens, an image model its price per image and per reference image, and a search, fetch, research, or transcription model its flat price per call.
Response headers
| Header | What it says |
|---|---|
X-Morse-Version | The running build. Every reply carries it. |
Errors
Every error has the same shape. Errors explains it and when to retry.
401The key is missing or wrong.3 codes
invalid_api_key- The Authorization header is missing or not Bearer, the key is unknown or revoked, a token that starts with sk_ is not a whole sk_morse_ key, or the bearer is no Morse key or login token at all, such as another service's API key (sk-proj-...).
api_key_expired- The key is past the expiry set on it. Make a new key, or set a later expiry before a key expires.
invalid_token- The bearer is shaped as a Morse login token (a JWT) but its signature fails or it has expired. An API key sends the whole sk_morse_ key.
403The key's account may not make this request.4 codes
account_disabled- Morse has disabled the key's account.
account_deleted- The key's account has been deleted.
account_under_18- The account's owner said they are under 18.
ip_not_allowed- The key allows requests only from its allowed IPs, and the request came from another address. The message names the address the proxy saw.
404No model or alias of that name is listed to this caller.1 code
model_not_found- The id names no model or alias that GET /v1/models lists to this caller, such as a model below the key's min_privacy or one only the Morse app may use.
429The account listed the models too often. Wait the number of seconds in Retry-After, then retry.1 code
rate_limit_exceeded- The account used up its 60 model lists a minute. All of the account's keys and its owner's use of the Morse app share them. Up to 7 lists can go at once, then one each second.
503Morse or the provider is busy or unavailable. Retry the request later.1 code
service_unavailable- Morse could not start the request. Retry it.