RadGenius Logo

RadGenius

Saturday, July 6, 2024

OpenAI Assistant Pricing (Updated for GPT-5.4 Mini)

James
OpenAI Assistant Pricing (Updated for GPT-5.4 Mini)

Adding an Assistant to your platform with RadGenius relies on the OpenAI API. Here’s a breakdown of the costs and how the GPT-5.4 Mini recommendation helps you balance quality, speed, and value.

RadGenius Assistants now recommend GPT-5.4 Mini by default. Internal helper tasks, such as title generation, image description, web-search synthesis, and assistant-instruction generation, use GPT-5.4 Nano where possible to keep those support calls fast and low cost.

If you need even more advanced reasoning or depth, you can upgrade any individual Assistant to a larger GPT-5 model such as GPT-5.5.

You choose the model per Assistant, giving you the flexibility to balance performance and cost.

GPT vs. Assistant: What’s the Difference?

GPTs are created on the ChatGPT website and require users to log in with an OpenAI account.

RadGenius Assistants are powered by the same underlying technology but are made to be embedded directly into your own website or app. No logins or subscriptions needed for your users, they chat right on your site.

What Does It Cost to Use a RadGenius Assistant?

When someone interacts with your Assistant, OpenAI charges a small fee based on how much text is processed (measured in tokens). You pay the costs directly to OpenAI, with no markup.

  • Good default for most embedded Assistants
  • Balances response quality, latency, and cost
  • Best fit when the assistant needs reliable general-purpose answers without using the largest model for every turn

GPT-5.5 (Optional Upgrade)

  • Best fit when the assistant needs deeper reasoning or higher-quality synthesis
  • More expensive than mini/nano models, so use it intentionally for Assistants where quality matters more than cost

OpenAI model prices can change, so check the OpenAI pricing page for current token rates before estimating production spend.

Tips to Keep Costs Low

Here are some ways to reduce costs without sacrificing quality:

1. Keep Messages Concise

Set clear expectations in your Assistant’s instructions:

  • Ask users for specific info
  • Tell the Assistant to be brief and to the point

Example instruction:
“Answer clearly in 1–2 sentences unless more detail is requested.”

2. Recognize Task Completion

Train your Assistant to say when it has answered the question.
This prevents conversations from dragging on unnecessarily.

Example reply:
“Looks like that covers it! Let me know if you need anything else.”

3. Monitor Usage

RadGenius includes simple analytics to show:

  • How often your Assistant is used
  • Conversation history with your users

Use this data to fine-tune your Assistant’s behavior and manage your costs as your traffic grows.

What About File Storage?

Every Assistant gets 1 GB of free file storage for uploading documents, PDFs, and other reference materials to its knowledge base.

To save space:

  • Convert large PDFs to plain text
  • Remove duplicate or outdated files regularly

With RadGenius, adding AI Assistants to your platform is affordable, allowing you to provide value without significant costs.