Skip to main content

OpenAI

OpenAI LLMs are powerful AI systems trained on massive amounts of text data, allowing them to generate text, translate languages, write different kinds of creative content, and answer your questions in an informative way.

Reasoning Models Support

OpenAI now supports reasoning models (O series models) with advanced reasoning capabilities!

Key Features:

  • Enable Reasoning: Allow the model to apply extra reasoning for complex tasks
  • ⚙️ Configurable Reasoning Effort: Choose from three levels:
    • Low: Basic reasoning with minimal latency impact
    • Medium: Balanced reasoning and performance
    • High: Maximum reasoning capability
  • ⚠️ Cost Impact: May slightly increase latency and cost (typically 20-50% depending on selected level)
  • 📋 Availability: Only applicable for O series models

Getting started with OpenAI LLMs

  1. Signup an OpenAI account.

  2. Choose your plan

  3. From the side navigation menu, choose API Keys and create a new API key.

    img OpenAI LLM img OpenAI LLM

Setting up the OpenAI LLM model in Unstract

Now that we have an API key from OpenAI, we can use it to set up an LLM profile on the Unstract platform. For this:

  • Sign into the Unstract Platform
  • From the side navigation menu, choose Settings 🞂 LLMs
  • Click on the New LLM Profile button
  • From the list of LLMs, choose OpenAI. You should see a dialog box where you enter details.
Reasoning Models Configuration

If you're using O series models, you'll have additional options:

Enable Reasoning

  • ☑️ Check this option to activate advanced reasoning capabilities
  • Only available for O series models (e.g., o1-preview, o1-mini)

Reasoning Effort Levels

  • 🔹 Low: Quick responses with basic reasoning
  • 🔹 Medium: Recommended for most use cases
  • 🔹 High: Maximum reasoning for complex tasks

⚠️ Important: Higher reasoning effort may increase response time and costs by 20-50%

img OpenAI LLM Configuration

  • For Name,enter a name for this connector.
  • In the API Key field, paste the API key we created on the OpenAI website.
  • For Model, refer to OpenAI model names
  • Leave the Max retries,Timeout and the API Base field to their default values.
  • For O series models: Enable reasoning and select your preferred reasoning effort level (Low/Medium/High)
  • Click on Test Connection and ensure it succeeds. You can finally click on Submit and that should create a new LLM Profile for use in your Unstract projects.