I Tested the New GPT-5.2—It Just Can't Compete With Google Gemini 3:
Despite OpenAI's bold claims of widespread improvements, GPT-5.2 feels largely the same as ...
pcmag.com/news/i-tested-the-new-gpt-52-it-just-cant-compete-with-google-gemini-3
Photo: pcmag.com · view article
Gemini 3 Outshines GPT-5.2 in Key AreasGoogle’s Gemini 3 excels in image generation and reasoning, while GPT-5.2 shows only minor improvements despite a price hike.
Key FactsDec 24, 2025 · Reading time: ~45 sec.
Summary created with AI assistance · Source: pcmag.com
Summary created with AI assistance · Source: pcmag.com
- OpenAI rushed GPT-5.2 to respond to Gemini 3’s launch, but its upgrades remain subtle and barely noticeable.
- Gemini 3’s Nano Banana Pro sets a new standard in AI image generation and delivers faster, more accurate complex reasoning.
- GPT-5.2 improves slightly in tasks like presentations and image recognition but does not match Gemini 3’s performance.
- Both models appeal to advanced users, yet Gemini 3 offers a clearly enhanced overall experience.
the biggest leap for GPT models in agentic coding since GPT-5— Jeff Wang, CEO of Windsurf, on GPT-5.2's capabilities and agentic coding
Story Details
| Last Updated | |
| Total Articles | 6 |
Main Topics
News on this topic
I Tried Vibe Coding With Different Gemini Models. Here's What I Learned: A slower "reasoning" model might do more of the work for you -- and keep vibe coding ...
cnet.com/tech/services-and-software/i-tried-vibe-coding-with-different-gemini-models-heres-what-i-learned/
Gemini could soon handle tasks in the background without holding your screen hostage
digitaltrends.com/computing/gemini-could-soon-handle-tasks-in-the-background-without-holding-your-screen-hostage/
Gemini on Android may soon process requests in the background: Gemini may soon get a new feature that allows the AI chatbot's overlay to process your ...
indianexpress.com/article/technology/tech-news-technology/google-gemini-android-overlay-background-process-query-10437168/
Similar to This
Previous
The New Stack Mar 15
A beginner’s guide to vibe coding: Vibe coding isn't just for experts — here's how to describe your way to a working app without writing a single line ...
thenewstack.io/beginners-guide-to-vibe-coding/
The New Stack Mar 05
OpenAI launches GPT-5.4 Thinking and Pro: The new model is smarter and makes fewer errors than its predecessor, but it comes at OpenAI's highest per-token ...
thenewstack.io/openai-launches-gpt-5-4/
I Vibe Coded an App With 3 Popular Chatbots. The Real Winner Is a Good Prompt: One project. Three chatbots. Which one wins overall?
cnet.com/tech/services-and-software/i-vibe-coded-an-app-with-3-popular-chatbots-the-real-winner-is-a-good-prompt/
I spent 7 hours with a vibe coding team at Google's Gemini hackathon: I followed a hackathon team as they raced to vibe code an app in seven hours at Google's ...
businessinsider.com/vibe-coding-team-embed-google-gemini-hackathon-singapore-2026-2
OpinionVibe coding lets non-coders set foot in a world previously inaccessible: Vibe coding is what happens when someone uses artificial intelligence (AI) tools like ...
financialexpress.com/opinion/vibe-codingnbsplets-non-coders-set-foot-in-a-world-previously-inaccessible/4106450/
Ziff Davis Jan 05
Perplexity AI vs. ChatGPT: Which One Is Better?: How does Perplexity compare to ChatGPT? We evaluate their performance and compare their features.
extremetech.com/computing/perplexity-ai-vs-chatgpt-which-one-is-better
Everything to know about Google Gemini’s most recent AI updates: Vibe coding, faster AI models, and real-time translation are some of the new updates from Google ...
euronews.com/next/2025/12/18/from-vibe-coding-to-faster-models-whats-new-in-googles-gemini-update
Z.ai debuts open source GLM-4.6V, a native tool-calling vision model
venturebeat.com/ai/z-ai-debuts-open-source-glm-4-6v-a-native-tool-calling-vision-model-for
OpenAI's ChatGPT 5.1 versus Google's Gemini 3: How they compare: After ChatGPT blindsided Google 2022 and turned OpenAI into the face of generative AI ...
businessinsider.com/openai-chatgpt-5-1-google-gemini-3-how-they-compare-2025-12
Mashable Dec 04 25
DeepSeek v3.2: What it is, how it compares to ChatGPT, how to try it: DeepSeek's latest AI models are here.
mashable.com/article/deepseek-v3-2-models-released
WIRED Nov 18 25
Gemini 3 Is Here—and Google Says It Will Make Search Smarter: Gemini 3 is skilled at reasoning, generating video, and writing code. Amid talk of an AI bubble ...
wired.com/story/google-launches-gemini-3-ai-bubble-search/
Google Jules Autonomous Coding Agent Exits Beta With New Pricing And Features: After over two months in the beta phase, Google's Jules Coding Agent is now ...
hothardware.com/news/googles-jules-ai-coding-agent-pricing-features
Deep Think is available now in the Google Gemini App: A version of the AI model achieved gold-medal standard at the International Mathematical Olympiad earlier ...
mashable.com/article/deep-think-google-gemini-app-available
Google Gemini deletes user’s code: ‘I have failed you completely and catastrophically’
mashable.com/article/google-gemini-deletes-users-code
Gemini Pro Review: My Go-To AI Chatbot Costs $20 Per Month: I tested Gemini Pro, and the Google chatbot is better than ChatGPT Plus.
cnet.com/tech/services-and-software/gemini-pro-review-my-go-to-ai-chatbot-costs-20-per-month/
New Atlas Jun 20 25
The hidden carbon cost of chatting to your AI: Study reveals AI models' CO2 emissions and accuracy trade-offs, urging users to choose energy-efficient ...
newatlas.com/ai-humanoids/cost-chatbot-environment/
Gemini’s new video analysis feature is here, but there’s more coming (APK teardown)
androidauthority.com/gemini-new-video-analysis-feature-3568561/
China's DeepSeek quietly releases upgraded R1 AI model, ramping up competition with OpenAI
cnbc.com/2025/05/29/chinas-deepseek-releases-upgraded-r1-ai-model-in-openai-competition.html
Key FactsDeepSeek Advances AI Model Amid U.S. Trade Restrictions
May 29, 2025 · Reading time: ~45 sec.
Summary created with AI assistance · Source: cnbc.com
Summary created with AI assistance · Source: cnbc.com
- Chinese startup DeepSeek quietly launched an enhanced version of its AI reasoning model R1 on Hugging Face.
- Its free, open-source R1 model gained attention for outperforming competitors like Meta and OpenAI with low costs and rapid development.
- According to LiveCodeBench metrics, the upgraded model closely trails OpenAI’s o4-mini and o3 reasoning models.
- Despite U.S. chip export controls, China continues to advance its AI capabilities, a fact acknowledged by Nvidia’s CEO Jensen Huang.
The assumption was always questionable, and now it's clearly wrong.— Jensen Huang, CEO of Nvidia, on U.S. export controls.
Next