Google Gemini, formerly known as Bard, is a generative artificial intelligence chatbot and virtual assistant developed by Google. It is powered by the family of large language models (LLMs) of the same name, after previously being based on LaMDA and PaLM 2. The Gemini architecture is trained natively on multiple data types, allowing the models to process and generate text, computer code, images, audio, and video simultaneously. Google distributes the technology in varying capacities, ranging from efficient on-device versions ("Nano") and cost-effective, high-throughput variants ("Flash") to high-compute models designed for complex reasoning ("Pro" and "Ultra"). The 1.5 and 3 model generations introduced extended context windows, enabling the analysis of large datasets such as entire codebases, long-form videos, or extensive document archives in a single prompt.
Gemini was first announced on December 6, 2023, and replaced existing Google branding for AI services. In February 2024, the Bard chatbot was renamed Gemini, and the "Duet AI" branding for Google Cloud and Workspace was retired in favor of the Gemini identifier. The models integrate into the Google ecosystem through the Gemini mobile app, which functions as an overlay assistant on Android devices, and through the Vertex AI platform for third-party developers.
History
Background
In November 2022, OpenAI launched ChatGPT, a chatbot based on the GPT-3 family of large language models. ChatGPT gained worldwide attention, becoming a viral Internet sensation. Alarmed by ChatGPT's potential threat to Google Search, Google executives issued a "code red" alert, reassigning several teams to assist in the company's artificial intelligence efforts. Sundar Pichai, the CEO of Google and parent company Alphabet, was widely reported to have issued the alert, but Pichai later denied this to The New York Times. In a rare move, Google co-founders Larry Page and Sergey Brin, who had stepped down from their roles as co-CEOs of Alphabet in 2019, attended emergency meetings with company executives to discuss Google's response to ChatGPT. Brin requested access to Google's code in February 2023, for the first time in years.
Google had unveiled LaMDA, a prototype LLM, earlier in 2021, but it was not released to the public. When asked by employees at an all-hands meeting whether LaMDA was a missed opportunity for Google to compete with ChatGPT, Pichai and Google AI chief Jeff Dean said that while the company's chatbot had similar capabilities to ChatGPT, there were risks to introducing an LLM that might spread false information, so they decided to wait. In January 2023, Google Brain's sister company DeepMind CEO Demis Hassabis hinted at plans for a ChatGPT rival, and Google employees were instructed to accelerate progress on a ChatGPT competitor, intensively testing "Apprentice Bard" and other chatbots. Pichai assured investors during Google's quarterly earnings investor call in February that the company had plans to expand LaMDA's availability and applications.
Bard
#### Announcement
On February 6, 2023, Google announced Bard, a generative artificial intelligence chatbot powered by LaMDA. Bard was first rolled out to a select group of 10,000 "trusted testers", before a wide release scheduled at the end of the month. The project was overseen by product lead Jack Krawczyk, who described the product as a "collaborative AI service" rather than a search engine, while Pichai detailed how Bard would be integrated into Google Search. Reuters calculated that adding ChatGPT-like features to Google Search could cost the company $6 billion in additional expenses by 2024, while research and consulting firm SemiAnalysis calculated that it would cost Google $3 billion. The technology was developed under the codename "Atlas", with the name "Bard" in reference to the Celtic term for a storyteller and chosen to "reflect the creative nature of the algorithm underneath".
Multiple media outlets and financial analysts described Google as "rushing" Bard's announcement to preempt rival Microsoft's planned February 7 event unveiling its partnership with OpenAI to integrate ChatGPT into its Bing search engine in the form of Bing AI (later rebranded as Microsoft Copilot), as well as to avoid playing "catch-up" to Microsoft. Microsoft CEO Satya Nadella told The Verge: "I want people to know that we made them dance." Tom Warren of The Verge and Davey Alba of Bloomberg News noted that this marked the beginning of another clash between the two Big Tech companies over "the future of search", after their six-year "truce" expired in 2021; Chris Stokel-Walker of The Guardian, Sara Morrison of Recode, and analyst Dan Ives of investment firm Wedbush Securities labeled this an AI arms race between the two.
After an "underwhelming" February 8 livestream in Paris showcasing Bard, Google's stock fell eight percent, equivalent to a $100 billion loss in market value, and the YouTube video of the livestream was made private. Many viewers also pointed out an error during the demo in which Bard gives inaccurate information about the James Webb Space Telescope in response to a query. Google employees criticized Pichai's "rushed" and "botched" announcement of Bard on Memegen, the company's internal forum, while Maggie Harrison of Futurism called the rollout "chaos". Pichai defended his actions by saying that Google had been "deeply working on AI for a long time", rejecting the notion that Bard's launch was a knee-jerk reaction.
One week after the Paris livestream, Pichai asked 80,000 employees to spend two to four hours on testing Bard internally, while Google executive Prabhakar Raghavan asked them to correct any errors Bard made. In the following weeks, Google employees criticized Bard in internal messages, citing safety and ethical concerns and calling on company leaders not to launch the service. Google executives launched the product, overruling a negative risk assessment report conducted by its AI ethics team. After Pichai suddenly laid off 12,000 employees later that month due to slowing revenue growth, remaining workers shared memes and snippets of their humorous exchanges with Bard soliciting its "opinion" on the layoffs.
Technical Architecture
Gemini models are built on a Transformer (architecture) architecture, similar to other Large language models, but with a key difference: they are trained natively on multiple data types. This multimodal training allows the models to process and generate text, code, images, audio, and video simultaneously, unlike earlier models that handled each modality separately. The architecture incorporates techniques such as Multi-Head Attention and Positional Encoding to handle long sequences. The 1.5 and 3 generations extended context windows, enabling the analysis of large datasets such as entire codebases, long-form videos, or extensive document archives in a single prompt. This is achieved through efficient attention mechanisms that reduce computational overhead.
Google distributes Gemini in several variants to balance performance and efficiency. The "Nano" variant is designed for on-device use, running efficiently on mobile hardware. The "Flash" variant is optimized for high-throughput, cost-effective tasks. The "Pro" variant handles complex reasoning, while the "Ultra" variant is the highest-compute model for the most demanding applications. These variants are integrated into Google's Google Cloud platform via Vertex AI, allowing third-party developers to access the models through APIs.
Integration and Availability
The Gemini mobile app functions as an overlay assistant on android devices, replacing the previous Google Assistant in some capacities. It is integrated into Google Workspace, including Gmail, Docs, and Sheets, under the Gemini branding. The models are also available through Google Cloud's Vertex AI platform, which provides tools for developers to build and deploy AI applications. In February 2024, the Bard chatbot was renamed Gemini, and the "Duet AI" branding for Google Cloud and Workspace was retired in favor of the Gemini identifier. This rebranding unified Google's AI offerings under a single name.
Reception and Controversies
The release of Gemini has generated technical praise and public controversy. Commentators have highlighted the models' benchmarks in coding and retrieval tasks as competitive with OpenAI's GPT-4 and GPT-5. However, the product launch faced criticism regarding the reliability of its outputs. In early 2024, Google suspended the model's ability to generate images of people after users reported historical inaccuracies and bias in its depictions of human subjects. Subsequent updates, including the Gemini 1.5 and 3 series released throughout 2025, focused on reducing hallucinations, improving latency, and enhancing agentic capabilities for autonomous research and software development. The controversies highlight ongoing challenges in Generative AI regarding bias and factual accuracy.