​
Overview

PromptL seamlessly integrates with OpenAI’s API. By default, PromptL formats prompts in the structure required by OpenAI, so you can use the output directly without additional processing.

​
Basic Example

Here’s how to generate a response from OpenAI using PromptL:

import { render } from 'promptl-ai'
import OpenAI from 'openai'

const prompt = `
---
model: gpt-4o
temperature: 0.6
---

Generate a joke about {{ topic }}.
`

const { messages, config } = await render({
  prompt,
  parameters: { topic: 'chickens' },
})

const client = new OpenAI({ apiKey: YOUR_OPENAI_API_KEY })
const response = await client.chat.completions.create({
  ...config,
  messages,
})

console.log(response.choices[0].message.content)

​
Key Features

  1. Default Adapter: PromptL automatically uses the OpenAI adapter for correct formatting.
  2. Role-Based Messages: OpenAI expects a role field (system, user, assistant) in messages, which PromptL handles for you.
  3. Configuration Pass-Through: Configuration options (e.g., temperature, model) are passed directly to OpenAI’s API.

​
Error Handling

When working with OpenAI, ensure you handle potential API errors gracefully:

try {
  const response = await client.chat.completions.create({
    ...config,
    messages,
  })
  console.log(response.choices[0].message.content)
} catch (error) {
  console.error('Error with OpenAI API:', error)
}

​
Next Steps

Use PromptL with your ProviderAnthropic