# Google Gemini 3 News Launch

Google Gemini 3 is a multimodal large language model family developed by Google DeepMind, released in November 2025 as the successor to Gemini 2.0. The Gemini 3 news launch was a major AI industry event, extending the Gemini line introduced in December 2023.

Google Gemini 3 is a family of multimodal large language models developed by [Google DeepMind](https://www.wikiprompt.org/wiki/google-deepmind) and released in November 2025. The Gemini 3 news launch refers to the public announcement of this release, which was presented as a major step forward in the [artificial intelligence](https://www.wikiprompt.org/wiki/artificial-intelligence) industry and as the next generation of the Gemini model line introduced in December 2023.

Gemini is the successor to Google's earlier LaMDA and PaLM 2 models and powers the Gemini chatbot. The name comes from the Gemini zodiac sign. The original Gemini family included Gemini Pro, Gemini Deep Think, Gemini Flash, and Gemini Flash Lite, and later versions expanded with Gemini 1.5 and Gemini 2.0. Gemini 3 was positioned as the most capable version yet, building on the multimodal approach that allows the model to process text, images, audio, video, and computer code.

## Background

Google first announced Gemini, a [large language model](https://www.wikiprompt.org/wiki/large-language-model) developed by its subsidiary Google DeepMind, during the Google I/O keynote on May 10, 2023. It was described as a more powerful successor to PaLM 2, which was also unveiled at the event. Google CEO Sundar Pichai said at the time that Gemini was still in its early developmental stages. Unlike many other language models, Gemini was designed to be multimodal from the start, meaning it could process multiple types of data simultaneously, including text, images, audio, video, and computer code. The project was a collaboration between DeepMind and Google Brain, two branches of Google that had been merged into Google DeepMind.

In an interview with Wired, DeepMind CEO Demis Hassabis said that Gemini's advanced capabilities would allow it to surpass OpenAI's ChatGPT, which runs on GPT-4. Hassabis highlighted the strengths of DeepMind's AlphaGo program, which gained worldwide attention in 2016 when it defeated Go champion Lee Sedol, and said that Gemini would combine the power of AlphaGo with other Google DeepMind language models. In August 2023, The Information reported that Google was targeting a launch date of late 2023 and hoped to surpass OpenAI and other competitors by combining conversational text capabilities with AI-powered image generation. Google co-founder Sergey Brin was reportedly summoned out of retirement to assist with Gemini's development, along with hundreds of engineers from Google Brain and DeepMind. Because Gemini was trained on transcripts of YouTube videos, lawyers were brought in to filter out potentially copyrighted material.

## Development before launch

On December 6, 2023, Pichai and Hassabis announced Gemini 1.0 at a virtual press conference. The initial release contained three models: Gemini Ultra for highly complex tasks, Gemini Pro for a wide range of tasks, and Gemini Nano for on-device tasks. At launch, Gemini Pro and Nano were integrated into Bard and the Pixel 8 Pro smartphone, respectively. Gemini Ultra was set to power a later "Bard Advanced" tier and become available to software developers in early 2024. Google stated that Gemini would not be made widely available until the following year because of the need for extensive safety testing. Gemini was trained on Google's Tensor Processing Units, and the name also referenced NASA's Project Gemini as well as the DeepMind-Google Brain merger.

Gemini Ultra was reported to have outperformed GPT-4, Anthropic's Claude 2, Inflection AI's Inflection-2, Meta's LLaMA 2, and xAI's Grok 1 on a variety of industry benchmarks. It was also the first language model to outperform human experts on the 57-subject Massive Multitask Language Understanding test, with a score of 90 percent. Gemini Pro was made available to Google Cloud customers on AI Studio and Vertex AI on December 13, 2023. In accordance with an executive order signed by U.S. President Joe Biden in October 2023, Google said it would share testing results of Gemini Ultra with the federal government. The company also engaged in discussions with the government of the United Kingdom to comply with principles from the AI Safety Summit at Bletchley Park.

In January 2024, Google partnered with [Samsung](https://www.wikiprompt.org/wiki/samsung-electronics) to integrate Gemini Nano and Gemini Pro into the Galaxy S24 smartphone lineup. The following month, Bard and Duet AI were unified under the Gemini brand, and Gemini Advanced with Ultra 1.0 was released through a new AI Premium tier of Google One. In February 2024, Google launched Gemini 1.5, describing it as more powerful than 1.0 Ultra, with a new architecture, a mixture-of-experts approach, and a larger one-million-token context window. The same month, Google debuted Gemma, a smaller, free, and open-source range of Gemini models. At the Google I/O keynote on May 14, 2024, Google announced Gemini 1.5 Flash and plans to integrate a desktop-optimized build of Gemini Nano into the Chrome browser through its Built-in AI architecture. Two updated models, Gemini-1.5-Pro-002 and Gemini-1.5-Flash-002, were released on September 24, 2024.

On December 11, 2024, Google announced Gemini 2.0 Flash Experimental. This version introduced a Multimodal Live API for real-time audio and video interactions, native image and controllable text-to-speech generation with watermarking, and integrated Google Search. In June 2025, Google introduced Gemini CLI, an open-source AI agent that brought Gemini capabilities to the terminal, offering coding, automation, and problem-solving features with generous free usage limits for individual developers. These releases set the stage for the Gemini 3 launch later that year.

## The November 2025 launch event

The Gemini 3 news launch took place in November 2025 as a virtual press event. Google DeepMind presented Gemini 3 as the successor to Gemini 2.0 and as the latest major release in the Gemini family. The event drew attention across the [generative AI](https://www.wikiprompt.org/wiki/generative-ai) industry, with comparisons to the original Gemini 1.0 announcement in December 2023. According to Google, Gemini 3 incorporated improvements in reasoning, long-context understanding, and multimodal output, building on features introduced in Gemini 2.0 such as native image generation and controllable text-to-speech.

As with earlier Gemini launches, the company emphasized safety testing and responsible deployment. Google said it would continue to share evaluation results with government agencies, following the precedent set with Gemini Ultra. The company also highlighted Gemini 3's integration with Google services, including Search, Chrome, and Google Cloud. At the event, Google announced that Gemini 3 would be available to developers through Vertex AI and AI Studio, extending the distribution model used for Gemini Pro in December 2023.

## Model family and technical architecture

Gemini 3 continued the multimodal design that distinguished the Gemini family from text-only large language models. The underlying architecture is based on the [transformer](https://www.wikiprompt.org/wiki/transformer) model, a [deep learning](https://www.wikiprompt.org/wiki/deep-learning) approach introduced in 2017 that has become standard in modern language models. Like Gemini 1.5, Gemini 3 was expected to use a mixture-of-experts design, in which different parts of the [neural network](https://www.wikiprompt.org/wiki/neural-network) are activated for different types of input. This approach helps the model handle large amounts of data while keeping computational costs manageable.

The Gemini family has historically included multiple tiers to serve different use cases. The original lineup comprised Gemini Ultra, Pro, and Nano, and later versions added Flash and Flash Lite. For Gemini 3, Google did not immediately disclose the full set of model variants, but the launch was expected to include Pro, Flash, and Deep Think tiers. As of the launch, detailed technical specifications such as parameter counts and training data sizes had not been fully released. Google stated that Gemini 3 was trained on its Tensor Processing Units, continuing the hardware strategy used for earlier Gemini models.

The model's multimodal capabilities allow it to process and generate text, images, audio, and video. This is a key differentiator in the [machine learning](https://www.wikiprompt.org/wiki/machine-learning) field, where many models are limited to a single modality. Gemini 3's ability to combine these modalities in a single system was presented as a step toward more general artificial intelligence systems that can interact with the world in richer ways.

## Industry and competitive context

The Gemini 3 launch came at a time of intense competition in the large language model market. [OpenAI](https://www.wikiprompt.org/wiki/openai) had released GPT-4 and continued to update its ChatGPT service, while [Anthropic](https://www.wikiprompt.org/wiki/anthropic) had developed the Claude series of models. Google positioned Gemini 3 as a direct competitor to these systems, building on the benchmark performance of earlier Gemini versions. Gemini Ultra had already outperformed GPT-4 and Claude 2 on several industry benchmarks, and Google claimed that Gemini 3 extended that lead, although independent verification was not yet available as of the launch.

The event also took place against a backdrop of growing investment in AI infrastructure. Major cloud providers, including [Google Cloud](https://www.wikiprompt.org/wiki/google-cloud), [Amazon Web Services](https://www.wikiprompt.org/wiki/amazon-web-services), and [Microsoft Azure](https://www.wikiprompt.org/wiki/azure), were competing to offer access to leading models. Google's decision to make Gemini 3 available through Vertex AI and AI Studio was part of a broader strategy to attract developers and enterprise customers. The company also faced pressure from open-source model releases, which led to the debut of Gemma in February 2024. Gemini 3's launch was seen as an attempt to maintain Google's position at the forefront of the AI industry while addressing concerns about safety, copyright, and the environmental impact of large-scale model training.

## Availability and ecosystem

Gemini 3 was made available through multiple channels at launch. Developers could access the model through Google Cloud's Vertex AI and AI Studio, following the same route used for Gemini Pro in December 2023. Google also said that Gemini 3 would be integrated into the Gemini chatbot and into other Google products, including Search, Chrome, and Workspace. This mirrored the integration plans announced for Gemini 1.0, which included Search, Ads, Chrome, Duet AI, and AlphaCode 2.

On the consumer side, Google had previously partnered with Samsung to bring Gemini Nano and Pro to the Galaxy S24, and Gemini 3 was expected to appear in future Android devices. The open-source Gemini CLI, introduced in June 2025, was also updated to support Gemini 3, giving developers a terminal-based interface for coding and automation tasks. Google's broader ecosystem strategy involved making Gemini available both through proprietary services and through open tools, a balance that the company had been refining since the release of Gemma.

## Reception and significance

The Gemini 3 news launch was widely reported as a major event in the AI industry. Analysts noted that the release came less than two years after the original Gemini 1.0 announcement, reflecting the rapid pace of development in the field. The event highlighted Google DeepMind's continued investment in multimodal AI and its efforts to compete with OpenAI and Anthropic. As of the launch, independent benchmark results for Gemini 3 had not yet been published, and the model's long-term impact remained to be seen. Nevertheless, the release reinforced Google DeepMind's position as one of the leading research organizations in artificial intelligence, with a model family that had evolved from the early LaMDA and PaLM 2 systems into a broad platform for generative AI.

---
Source: https://www.wikiprompt.org/wiki/google-gemini-3-news-launch
License: CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/)
Last updated: 2026-09-12T16:25:43.824214+00:00