# Get model in organization

`GET /v1/organizations/{organization}/models/{model}`, operation `models.GetModelInOrganization`

Gets a model.

## Parameters

| Name | In | Required | Type | Description |
| --- | --- | --- | --- | --- |
| `organization` | path | Yes | `string` | The organization's ID. |
| `model` | path | Yes | `string` | The model's ID. |

## Responses

### 200

OK

| Field | Type | Required | Description |
| --- | --- | --- | --- |
| `alias` | `string` | Yes | What callers send as model in their requests, such as gpt-5 or fast; unique in the organization, or in the project for a project's model. |
| `availability` | `Availability` | Yes | For an organization's model: the projects it is shared with, which may adopt it. A project's model is shared with none. |
| `availability.allProjects` | `boolean` | Yes | Every project, including those created later. |
| `availability.projects` | `string[]` | Yes | Projects, projects/{project_id}, when not all of them; at most 500. Past that, make the resource available to every project. |
| `createTime` | `string (date-time)` | Yes | When the model was created. |
| `displayName` | `string` | Yes | Name shown in user interfaces. An adopted model shows its source's unless it has its own. |
| `formats` | `string[]` | Yes | The request formats the gateway serves, directly or by translation. Translation supports stateless generation; provider-specific features without an equivalent are refused. |
| `info` | `ModelInfo` | No | What kosem knows of the provider's model; absent when it does not know it. |
| `info.contextWindow` | `integer (int64)` | Yes | How many tokens of prompt and answer the model takes together. |
| `info.deprecated` | `boolean` | Yes | Whether the provider is retiring the model. |
| `info.displayName` | `string` | Yes | The model's name, such as GPT-5 mini. |
| `info.inputs` | `string[]` | Yes | What the model takes in: text, image, pdf, audio or video. |
| `info.maxInputTokens` | `integer (int64)` | Yes | How many tokens of prompt the model takes, when fewer than its context window; 0 otherwise. |
| `info.maxOutputTokens` | `integer (int64)` | Yes | How many tokens the model writes at most. |
| `info.pricing` | `ModelPricing` | No | The provider's list prices; absent when kosem does not know them. |
| `info.reasoning` | `boolean` | Yes | Whether the model reasons before it answers. |
| `info.reasoningEfforts` | `string[]` | Yes | The reasoning effort levels the model takes, such as low and high. |
| `info.structuredOutput` | `boolean` | Yes | Whether the model's answers can be held to a JSON schema. |
| `info.toolCalls` | `boolean` | Yes | Whether the model calls tools. |
| `name` | `string` | Yes | Resource name, organizations/{organization_id}/models/{model_id} or projects/{project_id}/models/{model_id}. |
| `nativeFormats` | `string[]` | Yes | Formats served directly by the provider, without translation. |
| `pricing` | `ModelOwnPricing` | No | The pricing given the model, as kosem knows no list prices for its provider's model; absent when it has none. Requests are priced at it, for usage and spend limits, while info.pricing is absent. An adopted model shows its source's. |
| `pricing.cacheRead` | `Money` | No | A million prompt tokens read from the provider's cache; the input price unless set. |
| `pricing.cacheWrite` | `Money` | No | A million prompt tokens written to the provider's cache; the input price unless set. |
| `pricing.input` | `Money` | No | A million prompt tokens. Required unless pricedLike is set. Read, the price that applies, which for pricedLike is the catalog's; absent once the catalog has none. |
| `pricing.output` | `Money` | No | A million tokens of the answer, including reasoning. Required unless pricedLike is set. |
| `pricing.pricedLike` | `CatalogModelRef` | No | A model of kosem's catalog with list prices, which this model is priced at as the catalog changes. Set it, or input and output, not both. |
| `provider` | `string` | Yes | The provider that requests go through: organizations/{organization_id}/providers/{provider_id}, of the model's organization; or, for a project's own model, projects/{project_id}/providers/{provider_id}, the project's own. An adopted model shows its source's. |
| `providerModel` | `string` | Yes | The provider's name for the model, such as gpt-5-mini, which replaces alias in requests sent to it. An adopted model shows its source's. |
| `providerType` | `string` | Yes | The type of the model's provider, such as providerTypes/openai, which never changes. Those who may call the model see it, though they may not see the provider. |
| `source` | `string` | Yes | For a model a project adopted: the organization's model, organizations/{organization_id}/models/{model_id}, which serves its requests; empty for any other. Set when the model is created, and never changed. |
| `state` | `string` | Yes | Whether a project's model serves requests: NEEDS_ATTENTION, serving nothing, while its source, or its provider, its organization's, is no longer shared with its project, or while a policy enforces llmgateway.projectProvidersDisabled and it goes through the project's own provider. An organization's model is ACTIVE. Values: `ACTIVE`, `NEEDS_ATTENTION`. |
| `stateReason` | `string` | Yes | Why the model needs attention, when it does. |
| `updateTime` | `string (date-time)` | Yes | When the model was last updated. |

### default

Error

Base URL: `https://api.kosem.ai`. See [API conventions](https://docs.kosem.ai/reference/api-conventions.md) for authentication, pagination and errors, and [https://docs.kosem.ai/openapi.yaml](https://docs.kosem.ai/openapi.yaml) for the full OpenAPI document.