Take our pop quiz for Gemini’s first anniversary
Happy birthday, Gemini! A year ago, we introduced Gemini 1.0 — and we’ve been pretty busy since. Over the past year, we’ve expanded the Gemini family of models, found creative ways to integrate Gemini capabilities into the products you love and enabled developers and organizations all over the world to use Gemini to build products, work smarter and grow their businesses.
To celebrate, we wanted to take a look back at just a few milestones from the Gemini journey so far. But instead of just telling you what we’ve been up to, we thought it would be more fun to test your knowledge with a quiz.
So, how much do you remember from the first year of our Gemini era?
-
1/15
What was an early placeholder title for Gemini when it was still in its project phase?
-
Titan is right! Titan is also the name of Saturn's largest moon — so we did stick with something of a star-studded theme.
Actually, the answer is Titan, which is also the name of Saturn’s largest moon. While we stayed in the stars, we pivoted to Gemini.
Titan is right! Titan is also the name of Saturn's largest moon — so we did stick with something of a star-studded theme.
Actually, the answer is Titan, which is also the name of Saturn’s largest moon. While we stayed in the stars, we pivoted to Gemini.
Titan is right! Titan is also the name of Saturn's largest moon — so we did stick with something of a star-studded theme.
Actually, the answer is Titan, which is also the name of Saturn’s largest moon. While we stayed in the stars, we pivoted to Gemini.
Titan is right! Titan is also the name of Saturn's largest moon — so we did stick with something of a star-studded theme.
Actually, the answer is Titan, which is also the name of Saturn’s largest moon. While we stayed in the stars, we pivoted to Gemini.
2/15
Since launch, we’ve advanced the model’s capabilities — launching a new generation with Gemini 1.5, which had a breakthrough in long context understanding, referring to the amount of information the model can process. At ____tokens, we offer the longest context window of any large-scale foundation model to date, which powers developer use cases that no other model can handle.
-
Two million is correct! Context windows are important because they help AI models recall information during a session — just wanted to give you some added, ahem, context.
Not quite — the answer is 2 million! Context windows are important because they help AI models recall information during a session — just wanted to give you some added, ahem, context.
Two million is correct! Context windows are important because they help AI models recall information during a session — just wanted to give you some added, ahem, context.
Not quite — the answer is 2 million! Context windows are important because they help AI models recall information during a session — just wanted to give you some added, ahem, context.
Two million is correct! Context windows are important because they help AI models recall information during a session — just wanted to give you some added, ahem, context.
Not quite — the answer is 2 million! Context windows are important because they help AI models recall information during a session — just wanted to give you some added, ahem, context.
Two million is correct! Context windows are important because they help AI models recall information during a session — just wanted to give you some added, ahem, context.
Not quite — the answer is 2 million! Context windows are important because they help AI models recall information during a session — just wanted to give you some added, ahem, context.
3/15
Which Gemini model is known for its on-device capabilities?
-
There’s no “yes” in Nano, but yes, that's the right answer! Gemini Nano is our smaller, more efficient model that powers Android features like Smart Reply, Summarize and more — and can even work without an internet connection.
Sorry, but no — as in (Na)no. The right answer is Gemini Nano, our smaller, more efficient model that powers Android features like Smart Reply, Summarize and more — and can even work without an internet connection.
There’s no “yes” in Nano, but yes, that's the right answer! Gemini Nano is our smaller, more efficient model that powers Android features like Smart Reply, Summarize and more — and can even work without an internet connection.
Sorry, but no — as in (Na)no. The right answer is Gemini Nano, our smaller, more efficient model that powers Android features like Smart Reply, Summarize and more — and can even work without an internet connection.
There’s no “yes” in Nano, but yes, that's the right answer! Gemini Nano is our smaller, more efficient model that powers Android features like Smart Reply, Summarize and more — and can even work without an internet connection.
Sorry, but no — as in (Na)no. The right answer is Gemini Nano, our smaller, more efficient model that powers Android features like Smart Reply, Summarize and more — and can even work without an internet connection.
There’s no “yes” in Nano, but yes, that's the right answer! Gemini Nano is our smaller, more efficient model that powers Android features like Smart Reply, Summarize and more — and can even work without an internet connection.
Sorry, but no — as in (Na)no. The right answer is Gemini Nano, our smaller, more efficient model that powers Android features like Smart Reply, Summarize and more — and can even work without an internet connection.
4/15
What is the word that means Gemini can reason and process data from text, images, videos, code and more?
-
Multiple rounds of applause! You got it right, the answer is multimodal.
No, sorry — multiple choice is hard, but not for Gemini. The correct answer is multimodal.
Multiple rounds of applause! You got it right, the answer is multimodal.
No, sorry — multiple choice is hard, but not for Gemini. The correct answer is multimodal.
Multiple rounds of applause! You got it right, the answer is multimodal.
No, sorry — multiple choice is hard, but not for Gemini. The correct answer is multimodal.
Multiple rounds of applause! You got it right, the answer is multimodal.
No, sorry — multiple choice is hard, but not for Gemini. The correct answer is multimodal.
5/15
There are various models of Gemini — of these models, which would you use to summarize a short movie?
-
You’re a Gemini pro! The right answer is Gemini 1.5 Pro. Last May, we announced that 1.5 Pro has a context window of 2 million tokens, which means it can recall more information and summarize larger amounts of data — like that short movie we mentioned. If this question felt a little tricky, it might be because we announced this 2 million token context window at the same time we introduced Gemini 1.5 Flash — which, true to its name, is optimized for speed and efficiency.
It’s OK, you can’t be a pro at everything! The right answer is Gemini 1.5 Pro. Last May, we announced that 1.5 Pro has a context window of 2 million tokens, which means it can recall more information and summarize larger amounts of data — like that short movie we mentioned. If this question felt a little tricky, it might be because we announced this 2 million token context window at the same time we introduced Gemini 1.5 Flash — which, true to its name, is optimized for speed and efficiency.
You’re a Gemini pro! The right answer is Gemini 1.5 Pro. Last May, we announced that 1.5 Pro has a context window of 2 million tokens, which means it can recall more information and summarize larger amounts of data — like that short movie we mentioned. If this question felt a little tricky, it might be because we announced this 2 million token context window at the same time we introduced Gemini 1.5 Flash — which, true to its name, is optimized for speed and efficiency.
It’s OK, you can’t be a pro at everything! The right answer is Gemini 1.5 Pro. Last May, we announced that 1.5 Pro has a context window of 2 million tokens, which means it can recall more information and summarize larger amounts of data — like that short movie we mentioned. If this question felt a little tricky, it might be because we announced this 2 million token context window at the same time we introduced Gemini 1.5 Flash — which, true to its name, is optimized for speed and efficiency.
You’re a Gemini pro! The right answer is Gemini 1.5 Pro. Last May, we announced that 1.5 Pro has a context window of 2 million tokens, which means it can recall more information and summarize larger amounts of data — like that short movie we mentioned. If this question felt a little tricky, it might be because we announced this 2 million token context window at the same time we introduced Gemini 1.5 Flash — which, true to its name, is optimized for speed and efficiency.
It’s OK, you can’t be a pro at everything! The right answer is Gemini 1.5 Pro. Last May, we announced that 1.5 Pro has a context window of 2 million tokens, which means it can recall more information and summarize larger amounts of data — like that short movie we mentioned. If this question felt a little tricky, it might be because we announced this 2 million token context window at the same time we introduced Gemini 1.5 Flash — which, true to its name, is optimized for speed and efficiency.
You’re a Gemini pro! The right answer is Gemini 1.5 Pro. Last May, we announced that 1.5 Pro has a context window of 2 million tokens, which means it can recall more information and summarize larger amounts of data — like that short movie we mentioned. If this question felt a little tricky, it might be because we announced this 2 million token context window at the same time we introduced Gemini 1.5 Flash — which, true to its name, is optimized for speed and efficiency.
It’s OK, you can’t be a pro at everything! The right answer is Gemini 1.5 Pro. Last May, we announced that 1.5 Pro has a context window of 2 million tokens, which means it can recall more information and summarize larger amounts of data — like that short movie we mentioned. If this question felt a little tricky, it might be because we announced this 2 million token context window at the same time we introduced Gemini 1.5 Flash — which, true to its name, is optimized for speed and efficiency.
6/15
Gemini is the most widely available AI assistant. How many languages is it available in?
-
Correct! The Gemini app is available in 45 languages reaching 200 countries and territories.
Not quite: The Gemini app is available in 45 languages reaching 200 countries and territories.
Correct! The Gemini app is available in 45 languages reaching 200 countries and territories.
Not quite: The Gemini app is available in 45 languages reaching 200 countries and territories.
Correct! The Gemini app is available in 45 languages reaching 200 countries and territories.
Not quite: The Gemini app is available in 45 languages reaching 200 countries and territories.
Correct! The Gemini app is available in 45 languages reaching 200 countries and territories.
Not quite: The Gemini app is available in 45 languages reaching 200 countries and territories.
7/15
AI Overviews in Search, built with a custom Gemini model, are available in more than 100 countries and territories. How many people do they now reach on a monthly basis?
-
The answer is indeed 1 billion! AI Overviews aren’t just available to a lot of people – they’re also making Search more helpful: People prefer using Search with AI Overviews, and they’re using them to ask longer, more complex questions.
The correct answer is actually 1 billion! We’ve seen that people prefer using Search with AI Overviews, and they find their search results more helpful. And people using AI Overviews ask longer, more complex questions and end up exploring a wider range of websites, too.
The answer is indeed 1 billion! AI Overviews aren’t just available to a lot of people – they’re also making Search more helpful: People prefer using Search with AI Overviews, and they’re using them to ask longer, more complex questions.
The correct answer is actually 1 billion! We’ve seen that people prefer using Search with AI Overviews, and they find their search results more helpful. And people using AI Overviews ask longer, more complex questions and end up exploring a wider range of websites, too.
The answer is indeed 1 billion! AI Overviews aren’t just available to a lot of people – they’re also making Search more helpful: People prefer using Search with AI Overviews, and they’re using them to ask longer, more complex questions.
The correct answer is actually 1 billion! We’ve seen that people prefer using Search with AI Overviews, and they find their search results more helpful. And people using AI Overviews ask longer, more complex questions and end up exploring a wider range of websites, too.
The answer is indeed 1 billion! AI Overviews aren’t just available to a lot of people – they’re also making Search more helpful: People prefer using Search with AI Overviews, and they’re using them to ask longer, more complex questions.
The correct answer is actually 1 billion! We’ve seen that people prefer using Search with AI Overviews, and they find their search results more helpful. And people using AI Overviews ask longer, more complex questions and end up exploring a wider range of websites, too.
8/15
Google for Workspace customers report they’re saving an average of 105 ____ per user, per week using Gemini.
-
That’s right, 105 minutes is the correct answer! Now what to do with all that extra time…
Hey, it’s OK you didn’t get this one, just take a minute…or take 105 of them if you want!
That’s right, 105 minutes is the correct answer! Now what to do with all that extra time…
Hey, it’s OK you didn’t get this one, just take a minute…or take 105 of them if you want!
That’s right, 105 minutes is the correct answer! Now what to do with all that extra time…
Hey, it’s OK you didn’t get this one, just take a minute…or take 105 of them if you want!
That’s right, 105 minutes is the correct answer! Now what to do with all that extra time…
Hey, it’s OK you didn’t get this one, just take a minute…or take 105 of them if you want!
9/15
What was the first Android smartphone to feature Gemini Nano?
-
You put the “pro” in Pixel 8 Pro! That's right, Pixel 8 Pro was the first Android smartphone to feature Gemini Nano — but it’s certainly not the last.
You didn’t get this one right, but you’re still a pro! Pixel 8 Pro was the first Android smartphone and first Pixel to run Gemini Nano — but it’s certainly not the last.
You put the “pro” in Pixel 8 Pro! That's right, Pixel 8 Pro was the first Android smartphone to feature Gemini Nano — but it’s certainly not the last.
You didn’t get this one right, but you’re still a pro! Pixel 8 Pro was the first Android smartphone and first Pixel to run Gemini Nano — but it’s certainly not the last.
You put the “pro” in Pixel 8 Pro! That's right, Pixel 8 Pro was the first Android smartphone to feature Gemini Nano — but it’s certainly not the last.
You didn’t get this one right, but you’re still a pro! Pixel 8 Pro was the first Android smartphone and first Pixel to run Gemini Nano — but it’s certainly not the last.
You put the “pro” in Pixel 8 Pro! That's right, Pixel 8 Pro was the first Android smartphone to feature Gemini Nano — but it’s certainly not the last.
You didn’t get this one right, but you’re still a pro! Pixel 8 Pro was the first Android smartphone and first Pixel to run Gemini Nano — but it’s certainly not the last.
10/15
This year, we brought Gemini models to NotebookLM, our tool that takes various data sources and becomes an expert resource on a given topic. What is the name of the feature we introduced where two AI hosts discuss your source content?
-
Listen up, you got this one right! The answer is Audio Overviews, which take the sources you’ve added to a Notebook and turn the content into an engaging conversation between two AI hosts.
Listen up…that’s not quite right. The answer is Audio Overviews, which take the sources you’ve added to a Notebook and turn the content into an engaging conversation between two AI hosts.
Listen up, you got this one right! The answer is Audio Overviews, which take the sources you’ve added to a Notebook and turn the content into an engaging conversation between two AI hosts.
Listen up…that’s not quite right. The answer is Audio Overviews, which take the sources you’ve added to a Notebook and turn the content into an engaging conversation between two AI hosts.
Listen up, you got this one right! The answer is Audio Overviews, which take the sources you’ve added to a Notebook and turn the content into an engaging conversation between two AI hosts.
Listen up…that’s not quite right. The answer is Audio Overviews, which take the sources you’ve added to a Notebook and turn the content into an engaging conversation between two AI hosts.
Listen up, you got this one right! The answer is Audio Overviews, which take the sources you’ve added to a Notebook and turn the content into an engaging conversation between two AI hosts.
Listen up…that’s not quite right. The answer is Audio Overviews, which take the sources you’ve added to a Notebook and turn the content into an engaging conversation between two AI hosts.
11/15
True or false: All seven of Google’s products with 2 billion users use Gemini in one way, shape or form to make them more helpful.
-
That's right — you knew this was true! Products like Maps, Search and Gmail all use Gemini.
False is…well, false. This statement is true! (Head to Google Search to check, if you want.) Products like Maps, Search and Gmail all use Gemini.
That's right — you knew this was true! Products like Maps, Search and Gmail all use Gemini.
False is…well, false. This statement is true! (Head to Google Search to check, if you want.) Products like Maps, Search and Gmail all use Gemini.
12/15
At I/O 2024, Sundar talked about Gemini's future abilities to "think" multiple steps ahead and work across different environments — like the web, virtual worlds and the real world — to get things done for you. How did he refer to this?
-
Yes! The correct answer is “as an AI agent.” “I think about them as intelligent systems that show reasoning, planning and memory,” Sundar said at I/O. Maybe these agents will come out from under cover soon…
The correct answer is “as an AI agent.” “I think about them as intelligent systems that show reasoning, planning and memory,” Sundar said at I/O. Maybe these agents will come out from under cover soon…
Yes! The correct answer is “as an AI agent.” “I think about them as intelligent systems that show reasoning, planning and memory,” Sundar said at I/O. Maybe these agents will come out from under cover soon…
The correct answer is “as an AI agent.” “I think about them as intelligent systems that show reasoning, planning and memory,” Sundar said at I/O. Maybe these agents will come out from under cover soon…
Yes! The correct answer is “as an AI agent.” “I think about them as intelligent systems that show reasoning, planning and memory,” Sundar said at I/O. Maybe these agents will come out from under cover soon…
The correct answer is “as an AI agent.” “I think about them as intelligent systems that show reasoning, planning and memory,” Sundar said at I/O. Maybe these agents will come out from under cover soon…
Yes! The correct answer is “as an AI agent.” “I think about them as intelligent systems that show reasoning, planning and memory,” Sundar said at I/O. Maybe these agents will come out from under cover soon…
The correct answer is “as an AI agent.” “I think about them as intelligent systems that show reasoning, planning and memory,” Sundar said at I/O. Maybe these agents will come out from under cover soon…
13/15
In August, we integrated Gemini into Android. Which of the following things can you ask Gemini to help you with?
-
Phew, Gemini certainly keeps busy — that’s right, it can do all of these things on Android!
The bad news: You didn’t get this one right. The good news: Gemini on Android can do all of these things.
Phew, Gemini certainly keeps busy — that’s right, it can do all of these things on Android!
The bad news: You didn’t get this one right. The good news: Gemini on Android can do all of these things.
Phew, Gemini certainly keeps busy — that’s right, it can do all of these things on Android!
The bad news: You didn’t get this one right. The good news: Gemini on Android can do all of these things.
Phew, Gemini certainly keeps busy — that’s right, it can do all of these things on Android!
The bad news: You didn’t get this one right. The good news: Gemini on Android can do all of these things.
14/15
When we announced Gemini in December 2023, we also unveiled what is used both to train Gemini models and serve products that run it. What was it?
-
You got it! Technically, today isn’t just Gemini’s one-year anniversary: On December 6, 2023 we also made Cloud TPU v5p available, which at the time was our most powerful, scalable and flexible AI accelerator (since then we’ve launched the next-generation TPU, Trillium). In case you were wondering, TPUs are custom-designed chips that specialize in AI computing tasks.
The correct answer is Cloud TPU v5p, which at the time was our most powerful, scalable and flexible AI accelerator (since then we’ve launched the next-generation TPU, Trillium). In case you were wondering, TPUs are custom-designed chips that specialize in AI computing tasks.
You got it! Technically, today isn’t just Gemini’s one-year anniversary: On December 6, 2023 we also made Cloud TPU v5p available, which at the time was our most powerful, scalable and flexible AI accelerator (since then we’ve launched the next-generation TPU, Trillium). In case you were wondering, TPUs are custom-designed chips that specialize in AI computing tasks.
The correct answer is Cloud TPU v5p, which at the time was our most powerful, scalable and flexible AI accelerator (since then we’ve launched the next-generation TPU, Trillium). In case you were wondering, TPUs are custom-designed chips that specialize in AI computing tasks.
You got it! Technically, today isn’t just Gemini’s one-year anniversary: On December 6, 2023 we also made Cloud TPU v5p available, which at the time was our most powerful, scalable and flexible AI accelerator (since then we’ve launched the next-generation TPU, Trillium). In case you were wondering, TPUs are custom-designed chips that specialize in AI computing tasks.
The correct answer is Cloud TPU v5p, which at the time was our most powerful, scalable and flexible AI accelerator (since then we’ve launched the next-generation TPU, Trillium). In case you were wondering, TPUs are custom-designed chips that specialize in AI computing tasks.
You got it! Technically, today isn’t just Gemini’s one-year anniversary: On December 6, 2023 we also made Cloud TPU v5p available, which at the time was our most powerful, scalable and flexible AI accelerator (since then we’ve launched the next-generation TPU, Trillium). In case you were wondering, TPUs are custom-designed chips that specialize in AI computing tasks.
The correct answer is Cloud TPU v5p, which at the time was our most powerful, scalable and flexible AI accelerator (since then we’ve launched the next-generation TPU, Trillium). In case you were wondering, TPUs are custom-designed chips that specialize in AI computing tasks.
15/15
We introduced Gemma, our family of lightweight, state-of-the-art open AI models, in February. In July, we released Gemma 2, a more efficient and higher-performing version. Over the year, what other models did we add to the Gemma family?
-
The correct answer is all of them. CodeGemma specializes in code completion and code generation tasks; RecurrentGemma is well-suited for text generation tasks and requires less memory and can generate long sequences faster than Gemma. And PaliGemma is a vision-language model, so it can do things like image and short video caption generation, visual question answering and object detection. Just yesterday, we also released PaliGemma 2!
The correct answer is all of them. CodeGemma specializes in code completion and code generation tasks; RecurrentGemma is well-suited for text generation tasks and requires less memory and can generate long sequences faster than Gemma. And PaliGemma is a vision-language model, so it can do things like image and short video caption generation, visual question answering and object detection. Just yesterday, we also released PaliGemma 2!
The correct answer is all of them. CodeGemma specializes in code completion and code generation tasks; RecurrentGemma is well-suited for text generation tasks and requires less memory and can generate long sequences faster than Gemma. And PaliGemma is a vision-language model, so it can do things like image and short video caption generation, visual question answering and object detection. Just yesterday, we also released PaliGemma 2!
The correct answer is all of them. CodeGemma specializes in code completion and code generation tasks; RecurrentGemma is well-suited for text generation tasks and requires less memory and can generate long sequences faster than Gemma. And PaliGemma is a vision-language model, so it can do things like image and short video caption generation, visual question answering and object detection. Just yesterday, we also released PaliGemma 2!
The correct answer is all of them. CodeGemma specializes in code completion and code generation tasks; RecurrentGemma is well-suited for text generation tasks and requires less memory and can generate long sequences faster than Gemma. And PaliGemma is a vision-language model, so it can do things like image and short video caption generation, visual question answering and object detection. Just yesterday, we also released PaliGemma 2!
The correct answer is all of them. CodeGemma specializes in code completion and code generation tasks; RecurrentGemma is well-suited for text generation tasks and requires less memory and can generate long sequences faster than Gemma. And PaliGemma is a vision-language model, so it can do things like image and short video caption generation, visual question answering and object detection. Just yesterday, we also released PaliGemma 2!
The correct answer is all of them. CodeGemma specializes in code completion and code generation tasks; RecurrentGemma is well-suited for text generation tasks and requires less memory and can generate long sequences faster than Gemma. And PaliGemma is a vision-language model, so it can do things like image and short video caption generation, visual question answering and object detection. Just yesterday, we also released PaliGemma 2!
The correct answer is all of them. CodeGemma specializes in code completion and code generation tasks; RecurrentGemma is well-suited for text generation tasks and requires less memory and can generate long sequences faster than Gemma. And PaliGemma is a vision-language model, so it can do things like image and short video caption generation, visual question answering and object detection. Just yesterday, we also released PaliGemma 2!
You have answered 0/15 questions.
Since you made it this far, and since the traditional gift for an AI model’s anniversary is, in fact, not a quiz, we dropped a little treat for you over in AI Studio.
-
Introducing Gemini 2.0: our new AI model for the agentic era
-
Try Deep Research and our new experimental model in Gemini, your AI assistant
-
5 ways our latest Gemini models are changing retail
-
Why over 140,000 developers took our five-day generative AI course
-
5 ways I'm handling the holidays with Gemini
-
How developers are using Gemini API
-