All articles tagged with #gemini nano
Originally Published 5 months ago — by 9to5Google
Google is gradually rolling out AI Weather Reports to Pixel 8 and Pixel 8a devices with the Gemini Nano update, providing brief weather summaries that are similar to those available on Pixel 9 series, though they offer limited additional insights.
Originally Published 1 year ago — by 9to5Google
An Android AICore update is rolling out to the Pixel 8 and Pixel 8a, enabling on-device GenAl features through Gemini Nano. The update, which appears in Developer options, requires overnight charging and Wi-Fi for full functionality. Initially limited by RAM, Gemini Nano is now available on more devices, including the base Galaxy S24, and powers features like Summarize in Pixel Recorder and Gboard Smart Reply.
Originally Published 1 year ago — by TechSpot
Google's Gemini Nano AI model, previously unavailable on the Pixel 8 due to RAM limitations, will now be offered as a developer option in the next software update. The company aims to have AI features like smart reply "RAM resident," permanently occupying memory for instant access. However, enabling Gemini Nano may impact system performance, raising questions about the trade-off for users.
Originally Published 1 year ago — by Ars Technica
Google's latest AI model, Gemini Nano, is causing concerns due to its significant RAM usage, leading to limitations on certain devices such as the smaller Pixel 8. While the Pixel 8 Pro with 12GB of RAM will have access to Gemini Nano as a user-facing feature, the Pixel 8 will only be able to access it as a developer option due to its 8GB of RAM. Google aims to keep AI models "RAM-resident" for features like smart reply, but this could lead to trade-offs in system memory. The debate over the necessity of such AI features and their impact on device performance continues.
Originally Published 1 year ago — by Android Authority
Google has confirmed that the Gemini Nano on-device AI model will be available to all Pixel 8 users through Developer Options, enabling features such as Smart Reply in Gboard and summaries in the Recorder app. This comes after initial concerns about hardware limitations, with the AI model set to be included in the next Pixel Feature Drop as a developer preview.
Originally Published 1 year ago — by 9to5Google
Google announces that the Pixel 8 will receive Gemini Nano in a developer preview, enabling features like Summarize in Recorder and Gboard Smart Reply, despite previous claims of hardware limitations. The update, part of the next Pixel Feature Drop in June, will also allow developers to use AICore for their own applications. The smaller RAM in the Pixel 8 compared to the Pixel 8 Pro has been addressed, ensuring that running large language models does not impact user experience.
Originally Published 1 year ago — by TechCrunch
Google is set to introduce on-device AI-powered features like recording summaries and smart replies on the Pixel 8, utilizing the Gemini Nano model for on-device processing. These features will be available in a developer preview with the next Pixel feature drop, expanding beyond the Pixel 8 Pro and Galaxy S24. Google aims to gather feedback and foster innovation by enabling more enthusiasts and developers to experiment with Gemini Nano, marking a shift from primarily cloud-based AI features. Additionally, the "Circle to Search" feature will be extended to more Pixel and Samsung devices, with an update to support instant translation of on-screen content.
Originally Published 1 year ago — by The Verge
Google has announced that the Pixel 8 will receive support for on-device AI through Gemini Nano in the next Pixel feature drop, following pressure from Pixel fans and Android experts. Despite previous claims of hardware limitations, the Pixel 8 will now be able to utilize on-device recorder summaries and smart replies in Gboard, similar to the Pixel 8 Pro. This decision comes after Samsung began rolling out AI features to its previous-generation flagship phones, utilizing both on- and off-device AI, prompting questions about Google's initial explanation for the absence of on-device AI on the Pixel 8.
Originally Published 1 year ago — by Ars Technica
Google's latest AI models, including the Gemini Nano, cannot run on the smaller Pixel 8 due to mysterious "hardware limitations," despite being available on the larger Pixel 8 Pro and Samsung S24 family. This discrepancy raises questions about the actual differences between the two Pixel 8 models and the capabilities of the Google Tensor SoC. The lack of clarity in communication at the Pixel 8 launch event has left consumers feeling misled, and the limitations on the smaller Pixel 8 after just five months are disappointing, especially considering Google's promise of seven years of OS updates for the new Pixels.
Originally Published 1 year ago — by Android Authority
Google revealed on #TheAndroidShow that the Gemini Nano AI model will not be coming to the base model Pixel 8 due to hardware limitations, leaving owners wondering if it will ever arrive. However, the AI is expected to be available on more high-end devices in the near future.
Originally Published 1 year ago — by 9to5Google
Google announced that Gemini Nano, an AI model, will not be available on the Pixel 8 due to hardware limitations, but it is being worked on for more high-end devices. The difference in RAM between the Pixel 8 and 8 Pro may be a factor. Gemini Nano powers features like Recorder’s Summarize capability on the Pixel 8 Pro, and Google should consider implementing a cloud-based summarization for older devices that want the highly-requested feature.
Originally Published 1 year ago — by 9to5Google
MediaTek has announced that its Dimensity 9300 and 8300 chipsets are now "optimized" for Gemini Nano support, Google's on-device AI model. This collaboration with Google aims to ensure effective and efficient integration of Gemini Nano on smartphones powered by these MediaTek chips. While this doesn't immediately make Gemini Nano available on these devices, it simplifies the support process, with an APK for developers and OEMs expected soon. Additionally, MediaTek will showcase on-device generative AI using Llama 2 LLM at MWC 2024, also powered by Dimensity 9300 and 8300.
Originally Published 1 year ago — by Android Central
MediaTek has optimized its Dimensity 9300 and 8300 chips to support Google's Gemini Nano model, enabling on-device AI features. The integration leverages MediaTek's NeuroPilot toolkit and the chips' APU, offering benefits such as seamless performance, greater privacy, and lower operation costs. This development expands access to valuable AI features beyond Qualcomm, Samsung, and Google chips, benefiting flagship and midrange smartphone owners. MediaTek and Google plan to release an APK for developers and manufacturers to bring on-device Gemini Nano support to Dimensity 9300 and 8300 devices.
Originally Published 1 year ago — by 9to5Google
Google's Gboard's Gemini Nano Smart Reply feature, which initially supported only WhatsApp, has quietly added support for Line and KakaoTalk on the Pixel 8 Pro, with promises of more messaging apps to come. This on-device AI model provides higher-quality smart replies with conversational awareness, saving time and computational resources by not needing to query the cloud. To enable, users need to download the 1GB Gemini Nano model and enable Aicore Persistent in the settings.
Originally Published 2 years ago — by The Verge
Google is bringing its Gemini AI model, a large language model (LLM), to the Pixel 8 Pro smartphone through a reduced version called Gemini Nano. The Pixel 8 Pro will feature Gemini Nano in the auto-summarization feature of the Recorder app and the Smart Reply function of the Gboard keyboard. These features will work offline, providing fast and native performance. Google plans to expand Gemini Nano to more apps next year and integrate it into Android through a system service called AICore. While currently limited to English and requiring high-end chips, Gemini Nano has the potential to enhance the capabilities of high-end Android devices in the future.