Latest news with #GoogleGeminiApp
TechCrunch
14-05-2025
- Business
- TechCrunch
Google's Gemini chatbot can now more easily analyze GitHub projects
Gemini, Google's AI-powered chatbot, can now connect to GitHub — for users subscribed to the $20-per-month Gemini Advanced plan, that is. As of Wednesday, Gemini Advanced customers can directly add a public or private codebase on GitHub to Gemini to allow the chatbot to generate and explain code, debug existing code, and more. Users can connect GitHub to Gemini by clicking the '+' button in the prompt bar, selecting 'import code,' and pasting a GitHub URL. Gemini Advanced now connects with @github, making it a more powerful coding assistant. Directly connect to public or private GitHub repos to generate/modify functions, explain complex code, ask questions about your codebase, debug and more. Click the + button in the prompt bar,… — Google Gemini App (@GeminiApp) May 14, 2025 A word of warning: AI models, including Google's, still struggle to code quality software. Code-generating AI tends to introduce security vulnerabilities and errors, owing to weaknesses in areas like the ability to understand programming logic. One recent evaluation of Devin, a popular AI coding tool, found that it could only complete three out of 20 programming tests. The new GitHub integration arrives just a few days after OpenAI launched a GitHub connector for ChatGPT deep research, a tool that searches across the web and other sources to compile thorough research reports on a topic. AI companies, including Google and OpenAI, are moving quickly to ship new capabilities in an effort to make their products stand out in an increasingly crowded field. Case in point, OpenAI on Tuesday brought SharePoint and Microsoft OneDrive connectors to ChatGPT deep research.
Forbes
31-03-2025
- Business
- Forbes
Google's Best Gemini AI Chatbot Is Now Free To Use, But There's A Catch
Following a surprise announcement, Google has made its new AI chatbot, Gemini 2.5 Pro Experimental, free to use, giving everyone access to the company's 'most intelligent AI model' without a paid subscription. But there's a catch — free users won't get the whole experience. Moreover, the free version is missing one significant feature. Just days after its initial release to paying Google Gemini Advanced users on March 25, Google announced that Gemini 2.5 Pro (experimental) is now available to all Gemini users, with or without a paid Google One AI Premium subscription. In a recent tweet from the Google Gemini App account, the company states: Gemini users will now see the new Gemini 2.5 Pro Experimental model offered alongside existing models in the Gemini app or at Just select Gemini 2.5 Pro to use it. Since its release, Gemini 2.5 Pro has been making waves, outperforming competitors in key areas. According to Google, Gemini 2.5 Pro Experimental 'leads common benchmarks by meaningful margins' thanks to its strong reasoning and code capabilities. These claims are backed up by chart-topping scores on the LMArena leaderboard, where human voters rated it above strong competition from leading competitors Grok 3 Preview and ChatGPT 4.5 Preview. Forbes contributor Janakiram MSV provides more detail about Gemini 2.5 Pro in his recent article, but the key takeaway is that it's significantly more powerful than the previous Gemini 2.0 version. I can provide better responses and generate far superior code than previous iterations. Google reserves its best AI tools for paying Gemini Advanced users, so free users may wonder what the catch is. As you might expect, the 'free' version of Gemini 2.5 Pro Experimental isn't quite as capable as the full version available to paying customers. While the model itself remains unchanged, free users are limited in terms of the level of access they have to it. In a follow-up comment from the official Google Gemini App account, the company explains that free users 'have rate limits on this model, which do not apply to Advanced users. Your sub also gets you a longer context window.' Rate limits are how the company restricts how often you can make requests to Gemini, while the size of the context window determines how much information the AI model can process in one go, i.e., the number and size of documents, images and video it can process in one go. Gemini 2.5 Pro comes with a large context window of 1 million tokens, roughly equivalent to the complete works of Shakespeare. Google plans to expand this to 2 million tokens soon, further enhancing its capabilities. Gemini Advanced users can therefore use Gemini 2.5 Pro Experimental more frequently and upload bigger documents for processing. Your particular needs will determine whether the paid upgrade is worthwhile. Furthermore, only the paying Gemini Advanced users can use the latest AI model with Canvas, the shared digital space where you can create and edit documents and code interactively with Gemini. Much of the buzz around Gemini 2.5 Pro's capabilities depends on using Canvas for advanced AI-assisted tasks such as vibe coding, with several praising the tool's abilities to generate graphical apps that can run directly in your browser from a single text prompt. You'll still need Gemini Advanced to take full advantage of these capabilities. Google's free upgrade to Gemini 2.5 Pro Experimental brings a welcome surprise upgrade for all Gemini users, but you'll have to upgrade to Gemini Advanced to see what it can really do. However, it would be helpful if Google made the differences more explicit, as comments to Google's original announcement reveal that Gemini Advanced users are now questioning the need to retain their subscriptions. Google's API pricing for Gemini 2.5 Pro Experimental limits free users to just five requests per minute and 25 requests per day, whereas paying accounts get up to 20 requests per minute and 100 requests per day, along with double the maximum speed. Limits via the Gemini app and may vary. Follow @paul_monckton on Instagram.
