The Prompts API allows you to run AI prompts with your project. You can trigger prompts with different streaming options based on your application needs.

​

Available Endpoints

​

Run Prompt (Non-streaming)

Run a specific prompt and get the complete response in a single API call.

POST /prompt-runs/run-sync

​

Request Format

{
  "promptId": "string",
  "promptVersionId": "string",
  "replacements": {
    "key1": "value1",
    "key2": "value2"
  }
}

​

Response Format

Returns a complete JSON response

​

Run Prompt (HTTP/2 Streaming)

Run a specific prompt and stream the response using HTTP/2 protocol. This is useful for real-time applications where you want to display the response as it’s being generated.

POST /prompt-runs/async

​

Request Format

{
  "promptId": "string",
  "promptVersionId": "string",
  "replacements": {
    "key1": "value1",
    "key2": "value2"
  }
}

​

Response Format

Streams chunks of data as they become available.

​

Run Prompt (SSE Streaming)

Run a specific prompt and stream the response using Server-Sent Events (SSE) protocol. This is useful for web applications that need to display the response as it’s being generated.

POST /prompt-runs/run-sse

​

Request Format

{
  "promptId": "string",
  "promptVersionId": "string",
  "replacements": {
    "key1": "value1",
    "key2": "value2"
  }
}

​

Response Format

Streams events with data as they become available.

POST / prompt - runs / run - sync;
POST / prompt - runs / async;
POST / prompt - runs / run - sse;