Prompt Completion

Create Prompt Completion

POST /prompts/:id/completions

Creates a completion for the given prompt ID created in Portkey through the Prompts tab.

Creates a completion for the selected prompt ID

POST https://api.portkey.ai/v1/prompts/:id/completions

Path Parameters

Name	Type	Description
id*	String	The prompt ID to use for the completion request.

Request Body

Name	Type	Description
variables*	Object	The variables mentioned in the prompt template being used
hyperparameters	Multiple	Add any hyperparameters to the request to override the ones set in the prompt definition
stream	boolean	Incrementally stream the response using server-sent events. Defaults to false

200: OK

The API expects the prompt ID, prompt variables and optionally any hyperparameters to override for this request. It returns the chat completion or completion object in the unified format which is OpenAI signature compliant.

SDK Usage

The prompts.completions.create method in the Portkey SDK provides a way to generate a prompt completion.

Method Signature

NodeJS

prompt.completions.create(promptParams)

Python

prompts.completions.create(promptParams)

Parameters

• promptParams (Object): Parameters for the prompt completion request including promptID, variables and optionally any hyperparameters for the completion.

Example Usage

NodeJS

import Portkey from 'portkey-ai'

const portkey = new Portkey({
    apiKey: "PORTKEY_API_KEY",
})

// Make the prompt creation call with the variables
const promptCompletion = await portkey.prompts.completions.create({
    promptID: "Your Prompt ID",
    variables: {
       // The variables specified in the prompt
    }
})

// We can also override the hyperparameters
const promptCompletion = await portkey.prompts.completions.create({
    promptID: "Your Prompt ID",
    variables: {
       // The variables specified in the prompt
    },
    max_tokens: 250,
    presence_penalty: 0.2
})

Python

from portkey_ai import Portkey

client = Portkey(
    api_key="PORTKEY_API_KEY",  # defaults to os.environ.get("PORTKEY_API_KEY")
)

prompt_completion = client.prompts.completions.create(
    prompt_id="Your Prompt ID",
    variables={
       # The variables specified in the prompt
    }
)

print(prompt_completion)

# We can also override the hyperparameters
prompt_completion = client.prompts.completions.create(
    prompt_id="Your Prompt ID",
    variables={
       # The variables specified in the prompt
    },
    max_tokens=250,
    presence_penalty=0.2
)
print(prompt_completion)

REST

curl -X POST "https://api.portkey.ai/v1/prompts/PROMPT_ID/completions" \
-H "Content-Type: application/json" \
-H "x-portkey-api-key: $PORTKEY_API_KEY" \
-d '{
    "variables": {
        # variables to pass
    },
    "max_tokens": 250, # Optional
    "presence_penalty": 0.2 # Optional
}'

Streaming Mode

With the new /prompts/:id/completions route, it is mandatory to set stream:True while making the call—if you want to stream the responses. This is a departure from when you could toggle the stream mode in Portkey prompt playground UI, and have the request automatically return streamed response. Now, regardless of the stream toggle state in the prompt playground, if you want streaming mode, you must set it to true at the time of making the call.

NodeJS

const streamPromptCompletion = await portkey.prompts.completions.create({
    promptID: "Your Prompt ID",
    variables: {
       // The variables specified in the prompt
    },
    stream: true // Defaults to false
})

Python

stream_prompt_completion = client.prompts.completions.create(
    prompt_id="Your Prompt ID",
    variables={
       # The variables specified in the prompt
    },
    stream=True # Defaults to false
)

for chunk in stream_prompt_completion:
        print(chunk.choices[0].delta)

REST API

curl -X POST "https://api.portkey.ai/v1/prompts/PROMPT_ID/completions" \
-H "Content-Type: application/json" \
-H "x-portkey-api-key: $PORTKEY_API_KEY" \
-d '{
    "variables": {
        # variables to pass
    },
    "stream": true, # Defaults to false
    "max_tokens": 250, # Optional
    "presence_penalty": 0.2 # Optional
}'