> ## Documentation Index
> Fetch the complete documentation index at: https://www.voiceflow.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Prompts

API operations for prompts. Browse the routes below for parameters, request examples, and responses.

API version: Selected per operation

## Get many prompts

Reads several prompts by ID in one call. An ID that matches no prompt in the environment is reported instead of failing the request.

### Endpoint

```http
POST /v1/stable/prompt/get-many
```

Server: https://realtime-api.voiceflow.com

### Authentication

```json
[
  {
    "token": []
  }
]
```

`token`:

```json
{
  "scheme": "bearer",
  "bearerFormat": "JWT",
  "type": "http",
  "description": "Voiceflow bearer token"
}
```

### Query parameters

#### `projectID`

Required · string

The ID of the project to operate on.

```json
{
  "type": "string",
  "description": "The ID of the project to operate on."
}
```

#### `environmentAlias`

Required · string

The alias of the environment to operate on (e.g. `main`).

```json
{
  "type": "string",
  "description": "The alias of the environment to operate on (e.g. `main`)."
}
```

### Request body

Required

```json
{
  "required": true,
  "content": {
    "application/json": {
      "schema": {
        "$ref": "#/components/schemas/StablePromptGetManyRequest"
      }
    }
  }
}
```

### Responses

#### 200

Returns the prompts found, in request order, and `missing`, the requested IDs that match no prompt in this environment.

Schema:

```json
{
  "$ref": "#/components/schemas/StablePromptReadManyResponse"
}
```

Example:

```json
{
  "prompts": [
    {
      "id": "<string>",
      "name": "<string>",
      "system": "<string>",
      "settings": {
        "model": "<AIModel>",
        "maxTokens": 0,
        "temperature": 0,
        "responseFormat": "<string>",
        "reasoningEffort": "<AIReasoningEffort>",
        "serviceTier": "<AIParamsServiceTier>",
        "useStructuredOutput": true
      },
      "createdAt": "2026-09-10T00:00:00.000Z",
      "updatedAt": "2026-09-10T00:00:00.000Z",
      "createdByID": 0,
      "updatedByID": 0
    }
  ],
  "missing": [
    "<string>"
  ]
}
```

### Request examples

#### cURL

```bash
curl --request POST \
  --url 'https://realtime-api.voiceflow.com/v1/stable/prompt/get-many?projectID=<projectID>&environmentAlias=<environmentAlias>' \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
  "ids": [
    "<string>"
  ]
}'
```

#### JavaScript

```javascript
const response = await fetch('https://realtime-api.voiceflow.com/v1/stable/prompt/get-many?projectID=<projectID>&environmentAlias=<environmentAlias>', {
  method: 'POST',
  headers: {
    Authorization: 'Bearer <token>',
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    "ids": [
      "<string>"
    ]
  }),
});
const data = await response.json();
```

#### Python

```python
import requests

response = requests.post(
    'https://realtime-api.voiceflow.com/v1/stable/prompt/get-many?projectID=<projectID>&environmentAlias=<environmentAlias>',
    headers={'Authorization': 'Bearer <token>'},
    json={
      "ids": [
        "<string>"
      ]
    },
)
data = response.json()
```

#### Voiceflow CLI

```bash
vf prompt get-many --project-id <projectID> --environment-alias <environmentAlias> <<'JSON'
{
  "ids": [
    "<string>"
  ]
}
JSON
```


## Get prompt

Returns one prompt by ID. A prompt is the system text and model settings a workflow step in prompt mode sends to the model.

### Endpoint

```http
GET /v1/stable/prompt/{promptID}
```

Server: https://realtime-api.voiceflow.com

### Authentication

```json
[
  {
    "token": []
  }
]
```

`token`:

```json
{
  "scheme": "bearer",
  "bearerFormat": "JWT",
  "type": "http",
  "description": "Voiceflow bearer token"
}
```

### Path parameters

#### `promptID`

Required · string

The ID of the prompt to operate on.

```json
{
  "type": "string"
}
```

### Query parameters

#### `projectID`

Required · string

The ID of the project to operate on.

```json
{
  "type": "string",
  "description": "The ID of the project to operate on."
}
```

#### `environmentAlias`

Required · string

The alias of the environment to operate on (e.g. `main`).

```json
{
  "type": "string",
  "description": "The alias of the environment to operate on (e.g. `main`)."
}
```

### Responses

#### 200

Returns the prompt with its name, system text, settings (an empty object means the project defaults), timestamps, and the IDs of the users who created and last changed it.

Schema:

```json
{
  "$ref": "#/components/schemas/StablePromptReadResponse"
}
```

Example:

```json
{
  "prompt": {
    "id": "<string>",
    "name": "<string>",
    "system": "<string>",
    "settings": {
      "model": "<AIModel>",
      "maxTokens": 0,
      "temperature": 0,
      "responseFormat": "<string>",
      "reasoningEffort": "<AIReasoningEffort>",
      "serviceTier": "<AIParamsServiceTier>",
      "useStructuredOutput": true
    },
    "createdAt": "2026-09-10T00:00:00.000Z",
    "updatedAt": "2026-09-10T00:00:00.000Z",
    "createdByID": 0,
    "updatedByID": 0
  }
}
```

### Request examples

#### cURL

```bash
curl --request GET \
  --url 'https://realtime-api.voiceflow.com/v1/stable/prompt/<promptID>?projectID=<projectID>&environmentAlias=<environmentAlias>' \
  --header 'Authorization: Bearer <token>'
```

#### JavaScript

```javascript
const response = await fetch('https://realtime-api.voiceflow.com/v1/stable/prompt/<promptID>?projectID=<projectID>&environmentAlias=<environmentAlias>', {
  method: 'GET',
  headers: {
    Authorization: 'Bearer <token>',
  },
});
const data = await response.json();
```

#### Python

```python
import requests

response = requests.get(
    'https://realtime-api.voiceflow.com/v1/stable/prompt/<promptID>?projectID=<projectID>&environmentAlias=<environmentAlias>',
    headers={'Authorization': 'Bearer <token>'},
)
data = response.json()
```

#### Voiceflow CLI

```bash
vf prompt get --prompt-id <promptID> --project-id <projectID> --environment-alias <environmentAlias>
```


## Update prompt

Changes a prompt's name, its system text, or its whole settings object, and edits it in place so the step that uses it keeps pointing at the same ID. The `settings` value you send replaces the stored one, and an empty object returns it to the project defaults.

### Endpoint

```http
PATCH /v1/stable/prompt/{promptID}
```

Server: https://realtime-api.voiceflow.com

### Authentication

```json
[
  {
    "token": []
  }
]
```

`token`:

```json
{
  "scheme": "bearer",
  "bearerFormat": "JWT",
  "type": "http",
  "description": "Voiceflow bearer token"
}
```

### Path parameters

#### `promptID`

Required · string

The ID of the prompt to operate on.

```json
{
  "type": "string"
}
```

### Query parameters

#### `projectID`

Required · string

The ID of the project to operate on.

```json
{
  "type": "string",
  "description": "The ID of the project to operate on."
}
```

#### `environmentAlias`

Required · string

The alias of the environment to operate on (e.g. `main`).

```json
{
  "type": "string",
  "description": "The alias of the environment to operate on (e.g. `main`)."
}
```

### Request body

Required

```json
{
  "required": true,
  "content": {
    "application/json": {
      "schema": {
        "allOf": [
          {
            "$ref": "#/components/schemas/StablePromptUpdateRequest"
          }
        ]
      }
    }
  }
}
```

### Responses

#### 200

Confirms the prompt was updated. The body is a message, with `unresolvedReferences` for any variable or reference in the new system text that did not resolve.

Schema:

```json
{
  "$ref": "#/components/schemas/StableTextUpdateResponse"
}
```

Example:

```json
{
  "message": "<string>",
  "unresolvedReferences": [
    {
      "path": "<string>",
      "token": "<string>",
      "reason": "unknown"
    }
  ]
}
```

### Request examples

#### cURL

```bash
curl --request PATCH \
  --url 'https://realtime-api.voiceflow.com/v1/stable/prompt/<promptID>?projectID=<projectID>&environmentAlias=<environmentAlias>' \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
  "name": "<string>",
  "system": "<string>",
  "settings": {
    "model": "<AIModel>",
    "maxTokens": 0,
    "temperature": 0,
    "responseFormat": "<string>",
    "reasoningEffort": "<AIReasoningEffort>",
    "serviceTier": "<AIParamsServiceTier>",
    "useStructuredOutput": true
  }
}'
```

#### JavaScript

```javascript
const response = await fetch('https://realtime-api.voiceflow.com/v1/stable/prompt/<promptID>?projectID=<projectID>&environmentAlias=<environmentAlias>', {
  method: 'PATCH',
  headers: {
    Authorization: 'Bearer <token>',
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    "name": "<string>",
    "system": "<string>",
    "settings": {
      "model": "<AIModel>",
      "maxTokens": 0,
      "temperature": 0,
      "responseFormat": "<string>",
      "reasoningEffort": "<AIReasoningEffort>",
      "serviceTier": "<AIParamsServiceTier>",
      "useStructuredOutput": true
    }
  }),
});
const data = await response.json();
```

#### Python

```python
import requests

response = requests.patch(
    'https://realtime-api.voiceflow.com/v1/stable/prompt/<promptID>?projectID=<projectID>&environmentAlias=<environmentAlias>',
    headers={'Authorization': 'Bearer <token>'},
    json={
      "name": "<string>",
      "system": "<string>",
      "settings": {
        "model": "<AIModel>",
        "maxTokens": 0,
        "temperature": 0,
        "responseFormat": "<string>",
        "reasoningEffort": "<AIReasoningEffort>",
        "serviceTier": "<AIParamsServiceTier>",
        "useStructuredOutput": True
      }
    },
)
data = response.json()
```

#### Voiceflow CLI

```bash
vf prompt update --prompt-id <promptID> --project-id <projectID> --environment-alias <environmentAlias> <<'JSON'
{
  "name": "<string>",
  "system": "<string>",
  "settings": {
    "model": "<AIModel>",
    "maxTokens": 0,
    "temperature": 0,
    "responseFormat": "<string>",
    "reasoningEffort": "<AIReasoningEffort>",
    "serviceTier": "<AIParamsServiceTier>",
    "useStructuredOutput": true
  }
}
JSON
```


## Delete prompt

Deletes a prompt and the messages it holds. The call does not check which workflow step or condition holds the ID, so confirm nothing uses it first. The response carries only a confirmation message, so refetch the prompts to see what is left.

### Endpoint

```http
DELETE /v1/stable/prompt/{promptID}
```

Server: https://realtime-api.voiceflow.com

### Authentication

```json
[
  {
    "token": []
  }
]
```

`token`:

```json
{
  "scheme": "bearer",
  "bearerFormat": "JWT",
  "type": "http",
  "description": "Voiceflow bearer token"
}
```

### Path parameters

#### `promptID`

Required · string

The ID of the prompt to operate on.

```json
{
  "type": "string"
}
```

### Query parameters

#### `projectID`

Required · string

The ID of the project to operate on.

```json
{
  "type": "string",
  "description": "The ID of the project to operate on."
}
```

#### `environmentAlias`

Required · string

The alias of the environment to operate on (e.g. `main`).

```json
{
  "type": "string",
  "description": "The alias of the environment to operate on (e.g. `main`)."
}
```

### Responses

#### 200

Confirms the prompt was deleted. The body is a message only, so any step that held the ID still points at a prompt that no longer exists.

Schema:

```json
{
  "$ref": "#/components/schemas/StableDeleteResponse"
}
```

Example:

```json
{
  "message": "<string>"
}
```

### Request examples

#### cURL

```bash
curl --request DELETE \
  --url 'https://realtime-api.voiceflow.com/v1/stable/prompt/<promptID>?projectID=<projectID>&environmentAlias=<environmentAlias>' \
  --header 'Authorization: Bearer <token>'
```

#### JavaScript

```javascript
const response = await fetch('https://realtime-api.voiceflow.com/v1/stable/prompt/<promptID>?projectID=<projectID>&environmentAlias=<environmentAlias>', {
  method: 'DELETE',
  headers: {
    Authorization: 'Bearer <token>',
  },
});
const data = await response.json();
```

#### Python

```python
import requests

response = requests.delete(
    'https://realtime-api.voiceflow.com/v1/stable/prompt/<promptID>?projectID=<projectID>&environmentAlias=<environmentAlias>',
    headers={'Authorization': 'Bearer <token>'},
)
data = response.json()
```

#### Voiceflow CLI

```bash
vf prompt delete --prompt-id <promptID> --project-id <projectID> --environment-alias <environmentAlias>
```


## Create prompt

Creates a prompt, which is a system message plus optional model settings that a workflow step in prompt mode sends to the model. The prompt belongs to nothing until a step holds its ID.

### Endpoint

```http
POST /v1/stable/prompt
```

Server: https://realtime-api.voiceflow.com

### Authentication

```json
[
  {
    "token": []
  }
]
```

`token`:

```json
{
  "scheme": "bearer",
  "bearerFormat": "JWT",
  "type": "http",
  "description": "Voiceflow bearer token"
}
```

### Query parameters

#### `projectID`

Required · string

The ID of the project to operate on.

```json
{
  "type": "string",
  "description": "The ID of the project to operate on."
}
```

#### `environmentAlias`

Required · string

The alias of the environment to operate on (e.g. `main`).

```json
{
  "type": "string",
  "description": "The alias of the environment to operate on (e.g. `main`)."
}
```

### Request body

Required

```json
{
  "required": true,
  "content": {
    "application/json": {
      "schema": {
        "$ref": "#/components/schemas/StablePromptCreateRequest"
      }
    }
  }
}
```

### Responses

#### 201

Returns the new prompt with its ID, name, system text, and settings, and `unresolvedReferences` for any variable or reference in the system text that did not resolve.

Schema:

```json
{
  "$ref": "#/components/schemas/StablePromptCreateResponse"
}
```

Example:

```json
{
  "prompt": {
    "id": "<string>",
    "name": "<string>",
    "system": "<string>",
    "settings": {
      "model": "<AIModel>",
      "maxTokens": 0,
      "temperature": 0,
      "responseFormat": "<string>",
      "reasoningEffort": "<AIReasoningEffort>",
      "serviceTier": "<AIParamsServiceTier>",
      "useStructuredOutput": true
    },
    "createdAt": "2026-09-10T00:00:00.000Z",
    "updatedAt": "2026-09-10T00:00:00.000Z",
    "createdByID": 0,
    "updatedByID": 0
  },
  "unresolvedReferences": [
    {
      "path": "<string>",
      "token": "<string>",
      "reason": "unknown"
    }
  ]
}
```

### Request examples

#### cURL

```bash
curl --request POST \
  --url 'https://realtime-api.voiceflow.com/v1/stable/prompt?projectID=<projectID>&environmentAlias=<environmentAlias>' \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
  "name": "<string>",
  "system": "<string>",
  "settings": {
    "model": "<AIModel>",
    "maxTokens": 0,
    "temperature": 0,
    "responseFormat": "<string>",
    "reasoningEffort": "<AIReasoningEffort>",
    "serviceTier": "<AIParamsServiceTier>",
    "useStructuredOutput": true
  }
}'
```

#### JavaScript

```javascript
const response = await fetch('https://realtime-api.voiceflow.com/v1/stable/prompt?projectID=<projectID>&environmentAlias=<environmentAlias>', {
  method: 'POST',
  headers: {
    Authorization: 'Bearer <token>',
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    "name": "<string>",
    "system": "<string>",
    "settings": {
      "model": "<AIModel>",
      "maxTokens": 0,
      "temperature": 0,
      "responseFormat": "<string>",
      "reasoningEffort": "<AIReasoningEffort>",
      "serviceTier": "<AIParamsServiceTier>",
      "useStructuredOutput": true
    }
  }),
});
const data = await response.json();
```

#### Python

```python
import requests

response = requests.post(
    'https://realtime-api.voiceflow.com/v1/stable/prompt?projectID=<projectID>&environmentAlias=<environmentAlias>',
    headers={'Authorization': 'Bearer <token>'},
    json={
      "name": "<string>",
      "system": "<string>",
      "settings": {
        "model": "<AIModel>",
        "maxTokens": 0,
        "temperature": 0,
        "responseFormat": "<string>",
        "reasoningEffort": "<AIReasoningEffort>",
        "serviceTier": "<AIParamsServiceTier>",
        "useStructuredOutput": True
      }
    },
)
data = response.json()
```

#### Voiceflow CLI

```bash
vf prompt create --project-id <projectID> --environment-alias <environmentAlias> <<'JSON'
{
  "name": "<string>",
  "system": "<string>",
  "settings": {
    "model": "<AIModel>",
    "maxTokens": 0,
    "temperature": 0,
    "responseFormat": "<string>",
    "reasoningEffort": "<AIReasoningEffort>",
    "serviceTier": "<AIParamsServiceTier>",
    "useStructuredOutput": true
  }
}
JSON
```


## Update prompt settings

Changes individual model settings of a prompt, such as the model or the temperature, without resending the whole settings object. A key set to null is rejected, so clearing a key goes through the prompt update call.

### Endpoint

```http
PATCH /v1/stable/prompt/{promptID}/settings
```

Server: https://realtime-api.voiceflow.com

### Authentication

```json
[
  {
    "token": []
  }
]
```

`token`:

```json
{
  "scheme": "bearer",
  "bearerFormat": "JWT",
  "type": "http",
  "description": "Voiceflow bearer token"
}
```

### Path parameters

#### `promptID`

Required · string

The ID of the prompt to operate on.

```json
{
  "type": "string"
}
```

### Query parameters

#### `projectID`

Required · string

The ID of the project to operate on.

```json
{
  "type": "string",
  "description": "The ID of the project to operate on."
}
```

#### `environmentAlias`

Required · string

The alias of the environment to operate on (e.g. `main`).

```json
{
  "type": "string",
  "description": "The alias of the environment to operate on (e.g. `main`)."
}
```

### Request body

The settings keys to set. Keys you leave out keep their stored value. null is not accepted; to clear a key, send the whole settings object through update.

Required

```json
{
  "required": true,
  "description": "The settings keys to set. Keys you leave out keep their stored value. null is not accepted; to clear a key, send the whole settings object through update.",
  "content": {
    "application/json": {
      "schema": {
        "allOf": [
          {
            "$ref": "#/components/schemas/StablePromptUpdateSettingsRequest"
          }
        ]
      }
    }
  }
}
```

### Responses

#### 200

Confirms the prompt settings were updated. The body is a message only, so refetch the prompt to see its settings as they now stand.

Schema:

```json
{
  "$ref": "#/components/schemas/StableUpdateResponse"
}
```

Example:

```json
{
  "message": "<string>"
}
```

### Request examples

#### cURL

```bash
curl --request PATCH \
  --url 'https://realtime-api.voiceflow.com/v1/stable/prompt/<promptID>/settings?projectID=<projectID>&environmentAlias=<environmentAlias>' \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
  "model": "gpt-3.5-turbo-1106",
  "maxTokens": 0,
  "temperature": 0,
  "responseFormat": "<string>",
  "reasoningEffort": "off",
  "serviceTier": "default",
  "useStructuredOutput": true
}'
```

#### JavaScript

```javascript
const response = await fetch('https://realtime-api.voiceflow.com/v1/stable/prompt/<promptID>/settings?projectID=<projectID>&environmentAlias=<environmentAlias>', {
  method: 'PATCH',
  headers: {
    Authorization: 'Bearer <token>',
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    "model": "gpt-3.5-turbo-1106",
    "maxTokens": 0,
    "temperature": 0,
    "responseFormat": "<string>",
    "reasoningEffort": "off",
    "serviceTier": "default",
    "useStructuredOutput": true
  }),
});
const data = await response.json();
```

#### Python

```python
import requests

response = requests.patch(
    'https://realtime-api.voiceflow.com/v1/stable/prompt/<promptID>/settings?projectID=<projectID>&environmentAlias=<environmentAlias>',
    headers={'Authorization': 'Bearer <token>'},
    json={
      "model": "gpt-3.5-turbo-1106",
      "maxTokens": 0,
      "temperature": 0,
      "responseFormat": "<string>",
      "reasoningEffort": "off",
      "serviceTier": "default",
      "useStructuredOutput": True
    },
)
data = response.json()
```

#### Voiceflow CLI

```bash
vf prompt update-settings --prompt-id <promptID> --project-id <projectID> --environment-alias <environmentAlias> <<'JSON'
{
  "model": "gpt-3.5-turbo-1106",
  "maxTokens": 0,
  "temperature": 0,
  "responseFormat": "<string>",
  "reasoningEffort": "off",
  "serviceTier": "default",
  "useStructuredOutput": true
}
JSON
```


## Schema definitions

### AIModel

```json
{
  "type": "string",
  "enum": [
    "gpt-3.5-turbo-1106",
    "gpt-3.5-turbo",
    "gpt-4",
    "gpt-4o",
    "gpt-4o-mini",
    "gpt-4.1-2025-04-14",
    "gpt-4.1-mini-2025-04-14",
    "gpt-4.1-nano-2025-04-14",
    "gpt-o3-mini",
    "o3-2025-04-16",
    "o4-mini-2025-04-16",
    "gpt-5",
    "gpt-5-mini",
    "gpt-5-nano",
    "gpt-5.2",
    "gpt-5.4",
    "gpt-5.4-mini",
    "gpt-5.5",
    "gpt-5.6-sol",
    "gpt-5.6-terra",
    "gpt-5.6-luna",
    "gpt-6-sol",
    "gpt-6-luna",
    "gpt-6.1-sol",
    "gpt-realtime",
    "gpt-live-1",
    "claude-4-opus",
    "claude-4-sonnet",
    "claude-4.5-sonnet",
    "claude-4.6-sonnet",
    "claude-4.5-haiku",
    "claude-4.5-opus",
    "claude-4.7-opus",
    "bedrock-claude-4-sonnet",
    "bedrock-claude-4.5-sonnet",
    "bedrock-claude-4.6-sonnet",
    "bedrock-claude-5-sonnet",
    "bedrock-claude-5.5-sonnet",
    "bedrock-claude-4.5-haiku",
    "bedrock-claude-5.5-haiku",
    "bedrock-claude-4.5-opus",
    "bedrock-claude-4.7-opus",
    "bedrock-claude-4.8-opus",
    "voiceflow-core-4.0",
    "voiceflow-core-4.1",
    "voiceflow-core-4.2",
    "voiceflow-flash-4.2",
    "gemini-2.5-pro",
    "gemini-2.5-flash",
    "gemini-live-2.5-flash",
    "gemini-3-flash",
    "gemini-3.5-flash",
    "gemini-3.1-pro",
    "gpt-oss-20b",
    "gpt-oss-120b",
    "llama-guard-4",
    "llama-3.1-instant",
    "llama-3.3-versatile",
    "qwen3.6-27b",
    "grok-4.6",
    "voiceflow-flash-4.1",
    "kimi-k3",
    "gemini-flash-2",
    "gpt-4-turbo",
    "claude-3.5-haiku",
    "claude-3-opus",
    "claude-3.7-sonnet",
    "claude-v2",
    "llama-3.2-1b-preview",
    "gemini-pro-1.5",
    "claude-3-haiku",
    "claude-3-sonnet",
    "text-davinci-003",
    "claude-v1",
    "claude-instant-v1",
    "deep-seek-r1-distill-llama-70B",
    "claude-3.5-sonnet"
  ],
  "x-enumNames": [
    "GPT_3_5_TURBO_1106",
    "GPT_3_5_TURBO",
    "GPT_4",
    "GPT_4_O",
    "GPT_4_O_MINI",
    "GPT_4_1",
    "GPT_4_1_MINI",
    "GPT_4_1_NANO",
    "GPT_O_3_MINI",
    "GPT_O_3",
    "GPT_O_4_MINI",
    "GPT_5",
    "GPT_5_MINI",
    "GPT_5_NANO",
    "GPT_5_2",
    "GPT_5_4",
    "GPT_5_4_MINI",
    "GPT_5_5",
    "GPT_5_6_SOL",
    "GPT_5_6_TERRA",
    "GPT_5_6_LUNA",
    "GPT_6_SOL",
    "GPT_6_LUNA",
    "GPT_6_1_SOL",
    "GPT_REALTIME",
    "GPT_LIVE_1",
    "CLAUDE_4_OPUS",
    "CLAUDE_4_SONNET",
    "CLAUDE_4_5_SONNET",
    "CLAUDE_4_6_SONNET",
    "CLAUDE_4_5_HAIKU",
    "CLAUDE_4_5_OPUS",
    "CLAUDE_4_7_OPUS",
    "BEDROCK_CLAUDE_4_SONNET",
    "BEDROCK_CLAUDE_4_5_SONNET",
    "BEDROCK_CLAUDE_4_6_SONNET",
    "BEDROCK_CLAUDE_5_SONNET",
    "BEDROCK_CLAUDE_5_5_SONNET",
    "BEDROCK_CLAUDE_4_5_HAIKU",
    "BEDROCK_CLAUDE_5_5_HAIKU",
    "BEDROCK_CLAUDE_4_5_OPUS",
    "BEDROCK_CLAUDE_4_7_OPUS",
    "BEDROCK_CLAUDE_4_8_OPUS",
    "VOICEFLOW_CORE_4_0",
    "VOICEFLOW_CORE_4_1",
    "VOICEFLOW_CORE_4_2",
    "VOICEFLOW_FLASH_4_2",
    "GEMINI_2_5_PRO",
    "GEMINI_2_5_FLASH",
    "GEMINI_LIVE_2_5_FLASH",
    "GEMINI_3_FLASH",
    "GEMINI_3_5_FLASH",
    "GEMINI_3_1_PRO",
    "GPT_OSS_20_B",
    "GPT_OSS_120_B",
    "LLAMA_GUARD_4",
    "LLAMA_3_1_INSTANT",
    "LLAMA_3_3_VERSATILE",
    "QWEN_3_6_27_B",
    "GROK_4_6",
    "VOICEFLOW_FLASH_4_1",
    "KIMI_K_3",
    "GEMINI_FLASH_2",
    "GPT_4_TURBO",
    "CLAUDE_3_5_HAIKU",
    "CLAUDE_3_OPUS",
    "CLAUDE_3_7_SONNET",
    "CLAUDE_V_2",
    "LLAMA_3_2_1_B_PREVIEW",
    "GEMINI_PRO_1_5",
    "CLAUDE_3_HAIKU",
    "CLAUDE_3_SONNET",
    "DA_VINCI_003",
    "CLAUDE_V_1",
    "CLAUDE_INSTANT_V_1",
    "DEEP_SEEK_R_1_DISTILL_LLAMA_70_B",
    "CLAUDE_3_5_SONNET"
  ]
}
```

### AIParamsServiceTier

```json
{
  "type": "string",
  "enum": [
    "default",
    "flex",
    "priority"
  ],
  "x-enumNames": [
    "DEFAULT",
    "FLEX",
    "PRIORITY"
  ]
}
```

### AIReasoningEffort

```json
{
  "type": "string",
  "enum": [
    "off",
    "adaptive",
    "minimal",
    "low",
    "medium",
    "high"
  ],
  "x-enumNames": [
    "OFF",
    "ADAPTIVE",
    "MINIMAL",
    "LOW",
    "MEDIUM",
    "HIGH"
  ]
}
```

### PromptSettings

```json
{
  "type": "object",
  "properties": {
    "model": {
      "allOf": [
        {
          "$ref": "#/components/schemas/AIModel"
        }
      ]
    },
    "maxTokens": {
      "type": "number"
    },
    "temperature": {
      "type": "number"
    },
    "responseFormat": {
      "type": "string"
    },
    "reasoningEffort": {
      "nullable": true,
      "allOf": [
        {
          "$ref": "#/components/schemas/AIReasoningEffort"
        }
      ]
    },
    "serviceTier": {
      "description": "The service tier to request from models that support it. `priority` is faster processing: faster responses at a higher credit cost. When unset, the deprecated project-level `priorityProcessing` setting applies.",
      "nullable": true,
      "allOf": [
        {
          "$ref": "#/components/schemas/AIParamsServiceTier"
        }
      ]
    },
    "useStructuredOutput": {
      "type": "boolean"
    }
  }
}
```

### StableDeleteResponse

```json
{
  "type": "object",
  "properties": {
    "message": {
      "type": "string",
      "description": "A human-readable message confirming the deletion."
    }
  },
  "required": [
    "message"
  ]
}
```

### StablePrompt

```json
{
  "type": "object",
  "properties": {
    "id": {
      "type": "string"
    },
    "name": {
      "type": "string"
    },
    "system": {
      "$ref": "#/components/schemas/VariableReferenceString"
    },
    "settings": {
      "$ref": "#/components/schemas/PromptSettings"
    },
    "createdAt": {
      "type": "string",
      "format": "date-time",
      "pattern": "^(?:(?:\\d\\d[2468][048]|\\d\\d[13579][26]|\\d\\d0[48]|[02468][048]00|[13579][26]00)-02-29|\\d{4}-(?:(?:0[13578]|1[02])-(?:0[1-9]|[12]\\d|3[01])|(?:0[469]|11)-(?:0[1-9]|[12]\\d|30)|(?:02)-(?:0[1-9]|1\\d|2[0-8])))T(?:(?:[01]\\d|2[0-3]):[0-5]\\d(?::[0-5]\\d(?:\\.\\d+)?)?(?:Z))$"
    },
    "updatedAt": {
      "type": "string",
      "format": "date-time",
      "pattern": "^(?:(?:\\d\\d[2468][048]|\\d\\d[13579][26]|\\d\\d0[48]|[02468][048]00|[13579][26]00)-02-29|\\d{4}-(?:(?:0[13578]|1[02])-(?:0[1-9]|[12]\\d|3[01])|(?:0[469]|11)-(?:0[1-9]|[12]\\d|30)|(?:02)-(?:0[1-9]|1\\d|2[0-8])))T(?:(?:[01]\\d|2[0-3]):[0-5]\\d(?::[0-5]\\d(?:\\.\\d+)?)?(?:Z))$"
    },
    "createdByID": {
      "type": "number",
      "description": "The user who created the prompt."
    },
    "updatedByID": {
      "type": "number",
      "description": "The user who last changed the prompt, its settings or its system text."
    }
  },
  "required": [
    "id",
    "name",
    "system",
    "settings",
    "createdAt",
    "updatedAt",
    "createdByID",
    "updatedByID"
  ]
}
```

### StablePromptCreateRequest

```json
{
  "type": "object",
  "properties": {
    "name": {
      "description": "Defaults to \"Untitled prompt\", as the editor names one.",
      "type": "string",
      "minLength": 1,
      "maxLength": 255
    },
    "system": {
      "$ref": "#/components/schemas/VariableReferenceString"
    },
    "settings": {
      "allOf": [
        {
          "$ref": "#/components/schemas/StablePromptSettingsRequest"
        }
      ]
    }
  },
  "required": [
    "system"
  ],
  "additionalProperties": false
}
```

### StablePromptCreateResponse

```json
{
  "type": "object",
  "properties": {
    "prompt": {
      "$ref": "#/components/schemas/StablePrompt"
    },
    "unresolvedReferences": {
      "description": "References this write kept as literal text instead of binding, each with the reason. The write still succeeded.",
      "type": "array",
      "items": {
        "$ref": "#/components/schemas/StableUnresolvedReference"
      }
    }
  },
  "required": [
    "prompt"
  ]
}
```

### StablePromptGetManyRequest

```json
{
  "type": "object",
  "properties": {
    "ids": {
      "minItems": 1,
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "Prompt IDs to read. Duplicates are ignored; at most 50 distinct IDs."
    }
  },
  "required": [
    "ids"
  ],
  "additionalProperties": false
}
```

### StablePromptReadManyResponse

```json
{
  "type": "object",
  "properties": {
    "prompts": {
      "type": "array",
      "items": {
        "$ref": "#/components/schemas/StablePrompt"
      },
      "description": "The prompts found, in request order."
    },
    "missing": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "Requested IDs that match no prompt in this environment, in request order."
    }
  },
  "required": [
    "prompts",
    "missing"
  ]
}
```

### StablePromptReadResponse

```json
{
  "type": "object",
  "properties": {
    "prompt": {
      "$ref": "#/components/schemas/StablePrompt"
    }
  },
  "required": [
    "prompt"
  ]
}
```

### StablePromptSettingsRequest

```json
{
  "type": "object",
  "properties": {
    "model": {
      "allOf": [
        {
          "$ref": "#/components/schemas/AIModel"
        }
      ]
    },
    "maxTokens": {
      "type": "number"
    },
    "temperature": {
      "type": "number"
    },
    "responseFormat": {
      "type": "string"
    },
    "reasoningEffort": {
      "nullable": true,
      "allOf": [
        {
          "$ref": "#/components/schemas/AIReasoningEffort"
        }
      ]
    },
    "serviceTier": {
      "description": "The service tier to request from models that support it. `priority` is faster processing: faster responses at a higher credit cost. When unset, the deprecated project-level `priorityProcessing` setting applies.",
      "nullable": true,
      "allOf": [
        {
          "$ref": "#/components/schemas/AIParamsServiceTier"
        }
      ]
    },
    "useStructuredOutput": {
      "type": "boolean"
    }
  },
  "additionalProperties": false
}
```

### StablePromptUpdateRequest

```json
{
  "type": "object",
  "properties": {
    "name": {
      "type": "string",
      "minLength": 1,
      "maxLength": 255
    },
    "system": {
      "allOf": [
        {
          "$ref": "#/components/schemas/VariableReferenceString"
        }
      ]
    },
    "settings": {
      "allOf": [
        {
          "$ref": "#/components/schemas/StablePromptSettingsRequest"
        }
      ]
    }
  },
  "additionalProperties": false
}
```

### StablePromptUpdateSettingsRequest

```json
{
  "type": "object",
  "properties": {
    "model": {
      "allOf": [
        {
          "$ref": "#/components/schemas/AIModel"
        }
      ]
    },
    "maxTokens": {
      "type": "number"
    },
    "temperature": {
      "type": "number"
    },
    "responseFormat": {
      "type": "string"
    },
    "reasoningEffort": {
      "allOf": [
        {
          "$ref": "#/components/schemas/AIReasoningEffort"
        }
      ]
    },
    "serviceTier": {
      "allOf": [
        {
          "$ref": "#/components/schemas/AIParamsServiceTier"
        }
      ]
    },
    "useStructuredOutput": {
      "type": "boolean"
    }
  },
  "additionalProperties": false,
  "description": "The settings keys to set. Keys you leave out keep their stored value. null is not accepted; to clear a key, send the whole settings object through update."
}
```

### StableTextUpdateResponse

```json
{
  "type": "object",
  "properties": {
    "message": {
      "type": "string",
      "description": "A human-readable message confirming the update."
    },
    "unresolvedReferences": {
      "description": "References this write kept as literal text instead of binding, each with the reason. The write still succeeded.",
      "type": "array",
      "items": {
        "$ref": "#/components/schemas/StableUnresolvedReference"
      }
    }
  },
  "required": [
    "message"
  ]
}
```

### StableUnresolvedReference

```json
{
  "type": "object",
  "properties": {
    "path": {
      "type": "string",
      "description": "Dotted path of the field that carried it, e.g. `knowledgeBaseTool.filters[0].defaultValue`."
    },
    "token": {
      "type": "string",
      "description": "The reference as written, without braces."
    },
    "reason": {
      "type": "string",
      "enum": [
        "unknown",
        "secret-not-allowed",
        "path-on-non-variable",
        "resource-id",
        "invalid-name",
        "dangling"
      ],
      "description": "`unknown`: no variable or entity of that name. `secret-not-allowed`: the name is a project secret and this field does not accept one. `path-on-non-variable`: only a variable has fields. `resource-id`: a bare resource id, not a name. `invalid-name`: outside the `^\\w{1,64}$` name grammar. `dangling`: the reference rendered as a bare id, because its row is gone or its name is outside the name grammar.",
      "x-enumNames": [
        "UNKNOWN",
        "SECRET_NOT_ALLOWED",
        "PATH_ON_NON_VARIABLE",
        "RESOURCE_ID",
        "INVALID_NAME",
        "DANGLING"
      ]
    }
  },
  "required": [
    "path",
    "token",
    "reason"
  ]
}
```

### StableUpdateResponse

```json
{
  "type": "object",
  "properties": {
    "message": {
      "type": "string",
      "description": "A human-readable message confirming the update."
    }
  },
  "required": [
    "message"
  ]
}
```

### VariableReferenceString

```json
{
  "type": "string",
  "description": "Plain text; markdown is not interpreted. {variable_name} inserts a variable or entity, {variable_name.path} a field of a variable. Escape a literal { or \\ with a backslash. A name matching neither is stored as literal text and reported in unresolvedReferences; secrets are not available here."
}
```
