As you return to the office after your summer holidays, we’ve got two new LLMs in Insight’s AI Chat for you to try: GPT 5 and Gemini Image. Both models have more advanced capabilities compared to their predecessors.
Head to AI Chat, choose one of these new models, and do some experimenting!
GPT 5
GPT 5 is now available in Insight’s AI Chat. Launched in early August, this is OpenAI’s latest model that offers deeper reasoning capabilities, greater contextual understanding, and richer detail in its responses.
How is GPT 5 different from previous models?
- It may take slightly longer to generate an answer, depending on how much reasoning it does.
- The writing style is different (apply an AI style in chat for more consistency).
- It provides more detailed answers.
- GPT 5 is actually a combination of large and small AI models. Based on your question, GPT chooses which model to use to generate a response.
Try it: Go to AI Chat and select GPT 5 as your LLM. Use your favorite prompts and compare the results with GPT 4.1.
Gemini Image
You can now use Google’s newest image generation model in Insight’s AI Chat. For those of you using AI to help create social media and marketing graphics for your books, this might be the game-changer you’ve been waiting for.
Officially called Gemini 2.5 Flash Image, this model launched on August 26. It generates images quickly and more consistently preserves details of images you work with—like book covers.
Try it:
- Go to AI Chat and select Gemini Image as your LLM
- Upload an image to your chat conversation (click the image icon next to the model selector)
- Enter your prompt
Gemini Image Sample Prompts
Use and modify these sample prompts below to get started. You can also refer to Google’s prompt guide for Gemini 2.5 Flash Image.
Work with your book covers (upload your book cover into AI Chat, then enter these prompts):
Using the provided book cover image, create a [platform] social post. Keep the cover large and centered, add the text "[tagline]" in a [font style] at the top and "[CTA]" at the bottom. Use a gradient background derived from the cover palette. Clean, minimalist layout with generous negative space.This is a book cover. Generate an image of this 3D book against a neutral light background, angled slightly to show the pages.Generate a 3D image of this book cover as a hardcover book. Show it standing upright, slightly angled to show the pages, on a wooden table in a brightly lit coffee shop.This is a book cover. Create an image of a person holding this book in the foreground with one hand, the book cover clearly visible and in focus. In the background, a large bright living room, with large windows overlooking a big city. The background is softly blurred to emphasize the book. The person’s hand is visible, wearing a watch and bracelet, and has painted red nails. The scene is indoors, with bright light.