Gemini is Google’s AI chatbot competitor to services like Meta AI and OpenAI‘s ChatGPT, but it’s more complicated than just opening a chat window and asking questions. Even without a membership to use sophisticated models, Google provides a wide range of models to select from. Flash, which recently received an upgrade, is the one that balances the portability of Nano with the performance of Pro.
The stable version of the Gemini 2.0 Flash is now being made available to all users by Google. The experimental preview of 2.0 Flash, which was first made available in December 2024, will be replaced by the artificial intelligence (AI) model, which was revealed on Thursday. All users may access the new AI model through the mobile apps as well as the online client. The tech giant said that the new models will have enhanced capabilities, such as native support for audio and picture production, when it originally unveiled the Gemini 2.0 family of AI models. It is now available to both the online client and the mobile apps that will support the Gemini 2.0 Flash AI model.
Google said on Thursday that their Gemini 2.0 Flash AI model will be included in the Gemini app. The business stated in a statement that the updated model “provides everyday help with tasks like brainstorming, learning, or writing, delivering fast responses and stronger performance across a number of key benchmarks.”
According to Google, Flash is the Gemini model that you would use for everyday chores, including giving Gemini directions, which has replaced Google Assistant. According to the business, users may now have discussions with the AI via the Gemini app for Android using Gemini 2.0 Flash. The new model replaced Gemini 1.5, and according to analyst @MaxWeinbach on X (previously Twitter), it performs noticeably better than Claude 3.6 Sonnet and OpenAI’s GPT-4o.
Flash 2.0 was initially published in December of last year as an experiment, but it is now stable and available to all users. The update will be accessible to all users and is being rolled out to Gemini’s mobile and web apps. Nonetheless, the following Gemini models are now available for query processing if you are a Gemini Advanced subscriber; Flash 2.0, Flash 1.5, Pro 1.5, Pro 1.5 with Deep Research & Experimental Advanced 2.0 possibly in “the next few weeks.” allowing users to finish their tasks before moving to the new model.
In particular, the business claims that Flash 2.0 responds faster to tasks like learning, brainstorming, and writing. You would be correct to assume that this is obscure or insufficient to assess what has changed, since it appears to be a minor update rather than a significant advancement. For more precise picture production needs, Gemini now additionally uses the most recent version of Imagen 3.
The business stated that it was “working quickly” to include Gemini 2.0 into its products when it originally unveiled it in December. It gave Gemini users access to an experimental version of Gemini Flash 2.0 at the time.
Google also announced on Thursday that the most recent iteration of its Imagen 3 AI text-to-picture converter is now used by Gemini’s image production capabilities. Google claims that the model “follows your instructions with greater accuracy” and “delivers richer details and textures.”
The technology giant from Mountain View said in a blog post that the Gemini app would soon incorporate the Gemini 2.0 Flash AI model for every user. It was formerly limited to the online version. The next generation model will be available to Gemini Advanced customers, but it will also be accessible to free users. As of right now, it is unknown if the free tier has a rate restriction.
Google declared that the one million token context window with the AI model will also be accessible to Gemini Advanced customers. According to the business, this is sufficient to handle 1,500 pages of uploaded files. Additionally, users who pay will have access to premium features including Deep Research, Gems, and more.
The tech giant said that the Gemini 2.0 Flash offers multimodal output, including steerable text-to-speech (TTS) multilingual audio and picture production with text. Furthermore, agentic functions are also included in the AI model. 2.0 Flash natively calls third-party functions that have been defined by the user through the API, as well as features linked to code execution and Google Search.
Google released benchmark ratings for the AI model’s performance based on internal testing. According to reports, it performs better than Gemini 1.5 Pro on the Graduate-Level Google-Proof Q&A (GPQA), Natural2Code, MATH, and Massive Multitask Language Understanding (MMLU) benchmarks.
Notably, the most recent iteration of the Imagen 3 AI model is now supported in Gemini’s picture production. For the upcoming weeks, Gemini 1.5 Flash (for free users) and 1.5 Pro (for advanced users) will also be accessible. In order to give users adequate time to complete ongoing discussions and switch to the new AI model, Google created this extension.