Gemini Advanced or ChatGPT Plus—Which One Should You Pay For?


OpenAI’s ChatGPT has become a household name when it comes to AI tools, but Google isn’t standing on the sidelines. With its shiny new Gemini brand and Gemini Advanced service, Google has unleashed a powerful contender to OpenAI’s flagship.

Google claims that Gemini Advanced (an evolution of its previous Bard chatbot) beats GPT-4—but does it? If you only have $20 a month to spare, which paid multimodal AI tool should you use?

Let’s get this out of the way: there’s no clear “best” here. What sets these chatbots apart isn’t necessarily which is objectively “smarter” but which features, niche capabilities and external integrations best align with your existing tools and workflow.

Whether one of these powerful tools is worth your $20 a month depends entirely on the tools you already use and the tasks you need help with, and we’re here to help you decide with the nitty-gritty details.

Image generation: realism vs. fexibility

Both Gemini and ChatGPT Plus can transform your words into visual representations. Unlike Stable Diffusion and other image generators, these two understand natural language prompts. Gemini’s current focus is on achieving realism, but it falls short of Google’s separate ImageFX model, which, even in beta, blew our minds and might eventually be implemented into Gemini Advanced. But those two can only generate square 1024×1024 images.

OpenAI’s Dall-E 3 trades away realism for greater flexibility. You can specify image dimensions (square, portrait, 16:9, etc), a boon if you need visuals tailored to specific websites or designs and want to skip manual cropping. But it has a characteristic cartoon style that makes Dall-E images easy to spot from miles away.

Cyberpunk futuristic artist performing for for a crowd on stage with the word “DECRYPT” in neon lights in the background. Gemini (left) vs ChatGPT Plus (right)

Picking one or the other depends on your goals. Craving near-perfect product photos for an online store might still require specialized tools like those offered by Adobe or Corel. But for whimsical illustrations or playful brainstorming visuals, either of these contenders could suffice, leaving room for your personal preference on style.

Voice convenience: home vs. on-the-go

Sometimes, you want to hear your AI’s responses, whether it’s to allow multitasking or simply to give you a break from staring at a screen. thanks to its connection to the Google ecosystem, Gemini offers a seamless read-aloud feature. However, ChatGPT Plus has its own advantage: a native mobile app allowing your conversations to take place practically anywhere. In general terms, OpenAI’s voice feels more human, but it’s only available via smartphone.

Choosing boils down to how you work. If your AI use mainly takes place at a desk where integration with other Google tools is handy, Gemini wins. But for those constantly on the move, having those chats narrated while keeping your phone in your pocket might sway you towards ChatGPT Plus.This may seem important for visually impaired people, but many prefer other text-to-speech (TTS) methods that play content at extremely high speeds because they find the cadence of a “natural voice” too slow. In those cases, both sites are compatible with TTS, so it’s a wash.

Need for speed

Let’s peek under the hood, so to speak. While their “brain power” (token context) seems comparable, there is a huge difference in speed: Gemini Ultra is extremely fast when compared against GPT-4 (or Anthropic’s Claude AI). With Gemini Ultra you get GPT-4 quality outputs with GPT-3.5 speeds.

For example: Decrypt used the prompt “Please take your time and write an essay about why cryptocurrencies may play an important role in shaping the future of economic transactions.” Gemini Advanced took 12.14 seconds to write the whole essay whereas GPT-4 required nearly a minute—53.13 seconds, to be exact. The prior version of OpenAI’s chatbot, GPT-3.5 Turbo, took 11.06 to write its essay.

Privacy Matters

One of the biggest broad concerns about AI is privacy: where the prompts and other information you provide are transmitted, kept, and accessed. ChatGPT keeps your chats for 30 days, using that to improve its responses—but privacy-minded folks might be wary. Google, on the other hand, stores yours for a whopping 18 months.

Gemini (left) vs ChatGPT Plus (right) store users' interactions for different amounts of time and use it to train their models
Gemini (left) vs ChatGPT Plus (right) store users’ interactions for different amounts of time and use it to train their models

Both services raise different privacy concerns, making this choice as much about your feelings on data as much as the features. Fortunately, both also have an option to delete chats —and an option to share them.

ChatGPT’s niche win: PDF analysis

If you live and breathe PDFs, this contest is clear-cut: ChatGPT Plus can delve into those documents to extract insights, answer your questions about the document, and generally save you time. Google, probably for legal or technical reasons, hasn’t integrated such features into Gemini. This could be a non-issue if your PDFs are short chunks of text that you can just copy and paste. But for those juggling client paperwork, tables or research studies, it could be a dealbreaker.

This isn’t necessarily a permanent setback for Gemini, but for now, those working heavily with PDFs have a compelling reason to opt for ChatGPT’s enhanced capabilities. Also, Claude AI can analyze PDFs for free—and its model is almost as good and accurate as GPT-4.

Search Wars

If the quality of search results woven into your chatbot sessions matters, Gemini’s native link to Google wins hands-down.

Gemini Advanced lets users get real-time responses, with the model fetching data from Google Search on the go. Another extremely useful feature is that Gemini Advanced has a button that lets people use Google Search to double-check all facts in an interaction. This minimizes the impact of the occasional hallucination and helps with source scraping and fact-checking.

ChatGPT leans on Bing, which, while improving, can’t really compete with Google’s dominance.

It is important to note that every single answer provided by a chatbot is prone to include hallucinations. That’s the nature of generative AI. If a model cannot hallucinate, then it would reproduce already existing information by default. Fact-checking is extremely important when interacting with AI models.

What ChatGPT Plus has That Gemini Advanced doesn’t

Open Ecosystem and Third-party integrations: ChatGPT Plus stands apart by integrating with third-party apps —both plugins and user-generated GPTs. This creates endless possibilities for users to add plugins that directly interact with their AI assistant, streamlining workflows and unlocking specialized functionalities. Switching to Google Gemini will feel like a downgrade if you are used to working with Canva or Zapier and integrating them with ChatGPT in your workflow.

Personalized (and profitable) conversations: The GPT store aims to reward creators, promising ongoing innovation and growth within the ChatGPT ecosystem. If you want to make money creating personalized chatbots or want to try specifically tailored conversations with GPTs that ensure your chatbot knows more about a specific topic, adopts a specific style, and can be more personalized than any other option… then don’t look anywhere else and pay for ChatGPT Plus.

Mid-Conversation Edits for Efficiency: With ChatGPT Plus, you can adjust a previous prompt even within a multi-part conversation. For example, imagine a 6-shot interaction with six commands and six replies. In ChatGPT the user may edit the fourth command, and ChatGPT would generate a new answer only considering all the context before that interaction. This saves a lot of effort and makes the session more efficient. With Google, users would need to start a new session if they realized they made a mistake in a previous command.

This feature from ChatGPT Plus saves precious tokens and avoids restarting entire sessions, optimizing the interaction and maintaining your creative flow.

What Gemini Advanced gas That ChatGPT Plus doesn’t

Bundled Cloud Storage & Extras: The included 2TB Google One subscription is a substantial saving for those already within the Google ecosystem. Similar plans cost around $10 monthly on iCloud, and $12 on Dropbox. For these users, Switching to Gemini Advanced effectively cuts costs as it would be just an additional $10 on average for a top-tier AI chatbot instead of the $20 that users would pay if they were using ChatGPT Plus alongside a cloud service.

Powered by Google for Accuracy: Gemini Advanced verifies its responses against the vast knowledge of Google Search in real-time. This means accurate and up-to-date information at your fingertips, backed by Google’s reliability and relevance. ChatGPT with Bing is not as accurate.

Instant Drafts & Style Refinement: Gemini Advanced lets you tailor response styles (formal, casual, elaborate) with the click of a button. No extra prompts are needed! For refining your work, Gemini also offers multiple drafts with varying phrasing or focus, letting you quickly compare and improve upon your outputs.

Google Apps as Your AI Playground: You can Dive into Google Docs, plan travels, tinker with YouTube, edit content, and collaborate with the help of Gemini Advanced’s direct integration. Plus, those new and powerful photo editing tools that you get in Google Photos (Magic Eraser to delete objects using AI, camouflage, lighting tools, HDR enhancements, and more) become even more powerful when your AI can help tweak image parameters.

A Verdict… sort of

If you already pay for ChatGPT Plus and use its features often, Gemini may not bring enough novelty to your use case to justify another $20 monthly. Likewise, if your needs are fairly modest (writing assistance, creative brainstorming), either of these services would fit the bill.

But if you want to start paying for either one, those extras will provide the deciding factor. 2TB of Cloud storage, search accuracy, the Google suite – do these tilt you firmly one way, or does the promise of OpenAI’s third-party app integration leave you wanting to experiment? The choice really does lie with you because when it comes to AI, “it depends” remains the honest answer.

Edited by Ryan Ozawa.

Stay on top of crypto news, get daily updates in your inbox.



Fonte: decrypt.co