- Google has updated its Gemini AI chatbot with the Gemini 1.5 Flash model, enhancing response quality, speed, and image understanding, and making it accessible to all users.
- The model now supports 32,000 tokens, enabling longer and more complex interactions, and will be available in 40 languages across 230 countries, with additional safety measures for teenage users.
OUR TAKE
The expansion of Gemini 1.5 Flash to all users marks a significant step towards democratising access to advanced AI tools. Allowing file uploads in the free version enhances functionality and utility. Citations for responses improve transparency and trust, addressing concerns around AI accuracy and reliability. Making the tool accessible to teenagers with safety measures is a forward-thinking move.
–Vicky Wu, BTW reporter
What happened
Google has unveiled an update to its Gemini AI chatbot, integrating the Gemini 1.5 Flash, a lightweight large language model designed to compete with rivals such as Anthropic and OpenAI. The company has made the model accessible to all users, regardless of their subscription level. With this update, users will benefit from across-the-board improvements in the quality and speed of Gemini AI’s responses, along with enhanced reasoning and image understanding capabilities.
Initially showcased at Google I/O 2024 in May, Gemini 1.5 Flash was initially available to select paying customers. Now, the model is being rolled out to the free version of the Gemini mobile application and the web interface at gemini.google.com. The model’s capacity has been significantly boosted, supporting 32,000 tokens, a fourfold increase compared to the original version launched in May. This enhancement enables the handling of longer and more complex prompts, delivering more sophisticated and accurate responses.
Starting Thursday, Gemini 1.5 Flash will be available on the web and mobile platforms in 40 languages across approximately 230 countries. Additionally, teenagers aged 13 and older can now use Gemini, with safety measures implemented in collaboration with child safety organisations.
Also read: Apple Maps goes web-based, challenging Google with new features
Also read: Zuckerberg aims to rival OpenAI, Google with new Llama AI model
Why it’s important
Amar Subramanya, vice president of engineering for Gemini Experiences, has revealed that users will soon be able to upload files to the free version of the Gemini AI chatbot, a feature previously exclusive to the paid tier. “You’ll be able to do things like upload your economics study guide and ask Gemini to create practice questions,” Subramanya, wrote in a blog post. “Gemini will also soon be able to analyse data files for you, allowing you to uncover insights and visualize them through charts and graphics.”
Another forthcoming feature will empower Gemini to analyse data files and generate visualisations and charts based on the contents. To address the issue of AI hallucinations – responses that are nonsensical or inaccurate – Gemini will now provide citations for all of its responses, linking to the source materials it utilises. If the model produces a questionable response, users will be able to explore the cited sources to verify the accuracy of the information. These citations will also be included when using the Gmail extension, allowing users to ask Gemini questions related to their email inboxes.
