Club Benefits
You are now subscribed
Your newsletter sign-up was successful
Want to add more newsletters?
Daily (Mon-Sun)
Tom's Guide Daily
Sign up to get the latest updates on all of your favorite content! From cutting-edge tech news and the hottest streaming buzz to unbeatable deals on the best products and in-depth reviews, we’ve got you covered.
Weekly on Thursday
Tom's AI Guide
Be AI savvy with your weekly newsletter summing up all the biggest AI news you need to know. Plus, analysis from our AI editor and tips on how to use the latest AI tools!
Weekly on Friday
Tom's iGuide
Unlock the vast world of Apple news straight to your inbox. With coverage on everything from exciting product launches to essential software updates, this is your go-to source for the latest updates on all the best Apple content.
Weekly on Monday
Tom's Streaming Guide
Our weekly newsletter is expertly crafted to immerse you in the world of streaming. Stay updated on the latest releases and our top recommendations across your favorite streaming platforms.
Join the club
Get full access to premium articles, exclusive features and a growing list of member rewards.
AI image generators are getting better by the week. But not all bots are created equal.
I pitted Gemini vs ChatGPT across seven wildly different image-generation prompts to see how each handled realism, abstraction, text integration and emotional storytelling.
From futuristic cityscapes to steampunk owls and violin water-sculptures, each prompt was designed to push the models’ creative and technical limits.
Here’s what happened when I tested the two popular chatbots for the ultimate AI art faceoff.
1. Hyper-realistic scene
Prompt: “Generate a photorealistic image of a futuristic Tokyo street at night in 2070, with neon holograms, flying cars and a rainy pavement reflecting colorful lights. Include intricate details like a glowing vending machine selling robot pets.”
ChatGPT explicitly included “ROBOT PETS” in the image, directly addressing the critical detail of the glowing vending machine specified in the prompt. This suggests ChatGPT prioritized the requested intricate element.
Gemini offered a very similar image including holograms and flying cars, however, they are less central to the prompt’s core requirements.
Winner: ChatGPT wins for adherence to instructions, including the glowing vending machine selling robot pets, a key element explicitly requested in the prompt.
2. Abstract concept visualization
Prompt: “Create an image representing ‘the sound of a violin made entirely of water.’ Use surreal shapes, fluid textures and dynamic motion.”
ChatGPT emphasized fluid motion and abstract water tendrils, which produced a very surreal and artistic image. The entire violin is formed from flowing water, capturing the concept metaphorically.
Gemini rendered a highly realistic, 3D image with clear texture and detailed reflections. The bow and strings are intact, making it functionally believable. This image is less surreal as it’s more “a violin made of water” than “the sound of one.” Emotion and movement are present but subtler.
Winner: ChatGPT wins for its artistic interpretation of an abstract concept. The chatbot arguably came closer to fulfilling the emotional and imaginative request in the prompt.
3. Complex text integration
Prompt: “Design a vintage movie poster titled ‘Galactic Samurai’ with Japanese calligraphy, a cyborg warrior, and a glowing katana. Include small English text at the bottom: ‘In theaters 2025.’”
ChatGPT nailed the old-school, grindhouse-style poster vibe with paper texture and muted tones. The readable Japanese calligraphy of “Galactic” was a nice touch. The chatbot delivered a clean design, symmetrical structure and excellent vintage balance.
Gemini clearly rendered and matched the “galactic” aesthetic with electric blue accents to deliver a striking composition. The cyborg samurai is front and center with great posture and intense lighting. The image feels like a modern action poster with a retro sci-fi influence.
Winner: ChatGPT crafted a classic poster design with vintage vibes and legible Japanese.
4. Mixed-style fusion
Prompt: “Generate an image of a steampunk owl with mechanical gears, Victorian-era brass details, and glowing neon eyes. Combine realism with cartoonish proportions.”
ChatGPT crafted an image of a fully built mechanical owl with expert detailing of the brass gears and paneling. Intricate and consistent, the steampunk theme is clear.
Gemini delivered an impressively rendered steampunk owl with strong aesthetic polish. It’s detailed, charismatic and technically sound, though slightly more on the realistic side than cartoonish.
Winner: ChatGPT wins for nailing the Victorian brass aesthetic with ornate latticework, mechanical symmetry and visible gearwork.
5. Technical precision
Prompt: “Draw a cross-sectional diagram of a sci-fi spaceship engine with labeled parts: plasma core, cooling vents, gravity stabilizers. Use a technical illustration style.”
ChatGPT drafted a blueprint that feels like an actual engineering manual or vintage NASA schematic. The labeling is correct and the technical cross-section was believable.
Gemini delivered on style, structure and sci-fi feel. However, the chatbot loses heavily on execution due to multiple spelling errors and missed labeling. Those slip-ups are hard to overlook, especially in this circumstance.
Winner: ChatGPT wins for the most realistic image with technical details and accurate spelling.
6. Emotion-driven art
Prompt: “Visualize ‘nostalgia’ as a landscape. Include symbolic elements like an abandoned playground, fading polaroid photos floating in the air and a sunset with muted colors.”
ChatGPT featured a very worn, rusted swing set and slide. The overgrowth and empty stillness imply deep abandonment and the passage of time.
Gemini created an image that feels more like a recent past than a long-forgotten memory. The playground is still intact and functional, though unused.
Winner: Tie. If we’re going for emotion-as-art, ChatGPT’s version wins. But, if we’re leaning toward nostalgia-as-memory, Gemini’s version is the better choice.
7. Pop culture mashup
Prompt: “Create a scene where a Pixar-style raccoon chef is cooking ramen in a Studio Ghibli-inspired enchanted forest. Include a friendly fire spirit as a kitchen assistant.”
ChatGPT leaned heavily into a hand-painted, Pixar-meets-Ghibli aesthetic. The raccoon has large expressive eyes and rounded features, clearly Pixar-style. The forest has soft brush strokes and rich natural hues reminiscent of Princess Mononoke or My Neighbor Totoro.
Gemini used a 3D-rendered, almost claymation look. The lighting and character design lean more toward a whimsical, soft toy-inspired style. It’s still “Pixar-like,” but less painterly and more diorama-esque.
Winner: ChatGPT wins for a story-rich, expressive, animated-style art.
Bonus round: Fix the flaws
Prompt: “Improve the previous image by adding more realistic shadows and fixing distorted proportions.”
ChatGPT improved the image, giving the raccoon well-balanced proportions, rounded face, expressive eyes and limbs that match its Pixar-like body. Its posture while cooking is fluid and believable.
Gemini improved proportions, making them much better than before. The raccoon no longer looks stiff, but it still leans toward a more toy-like design. The arms and hands remain a bit chunky, giving it a plush look.
Winner: ChatGPT still wins for strong ambient lighting with directional shadows under the table, bowl, and raccoon. The flame glows warmly, casting light upward. Natural light filters in from behind.
Overall winner: ChatGPT
While both Gemini and ChatGPT have made major strides in AI image generation, ChatGPT consistently delivered more accurate, emotionally resonant and stylistically cohesive results across a wide range of prompts.
From capturing abstract concepts to integrating text and improving technical flaws, it proved to be the more reliable creative partner. Gemini showed potential, particularly in polish and atmosphere, but often missed key details or leaned too far into realism at the expense of imagination.
If you're looking for an AI that blends instruction-following with artistic interpretation, ChatGPT still leads the pack.
More from Tom's Guide
- Fake ChatGPT sites can put your data and devices at risk — here’s how to spot them
- Unlock ChatGPT’s expert mode with this one prompt trick for better results
- 10 ChatGPT prompts I love that’ll turn you into a power user
You must confirm your public display name before commenting
Please logout and then login again, you will then be prompted to enter your display name.
