Gemini Models

Technologies

Updated production-ready Gemini models, reduced 1.5 Pro pricing, increased rate limits, and more

The most general and capable AI models we've ever built.

Our most flexible models yet

Each Gemini model is built for its own set of use cases, making a versatile model family that runs efficiently on everything from data centers to on-device.

1.0

Ultra

Our largest model for highly complex tasks.

1.5

Pro

Our best model for general performance across a wide range of tasks.

1.5

Flash

Our lightweight model, optimized for speed and efficiency.

1.0

Nano

Our most efficient model for on-device tasks.

Project Astra

Project Astra explores the future of AI assistants. Building on our Gemini models, we’ve developed AI agents that can quickly process multimodal information, reason about the context you’re in, and respond to questions at a conversational pace, making interactions feel much more natural.

The demo shows two continuous takes: one with the prototype running on a Google Pixel phone and another on a prototype glasses device.

Natively multimodal

Gemini models are built from the ground up for multimodality, seamlessly combining and understanding text, code, images, audio, and video.

Following content is a visual/ descriptive representation of the functionality of Gemini:

Gemini models can generate code based on different kinds of inputs.

Gemini

I see a murmuration of starlings, so I coded a flocking simulation.

            class Boid {
              constructor(x, y) {
                this.pos = new p5.Vector(x, y);
                this.vel = p5.Vector.random2D();
                this.vel.setMag(random(2, 4));
                this.acc = new p5.Vector();
                this.maxForce = 0.2;
                this.maxSpeed = 4;
              }
          }

Gemini models can generate text and images, combined.

Could Gemini show me ideas for what to make?

Gemini

How about an octopus with blue and pink tentacles?

Gemini models can understand and perform tasks involving several different written languages.

Could Gemini explain what this means?

Gemini

I see the time signature is 6/8. This means there are 6 eighth notes in each measure.

The dynamic marking is piano, which means to play softly. Andante grazioso means to play at a graceful walking pace.

Longer context

1.5 Pro and 1.5 Flash both have a default context window of up to one million tokens — the longest context window of any large scale foundation model. They achieve near-perfect recall on long-context retrieval tasks across modalities, unlocking the ability to process long documents, thousands of lines of code, hours of audio, video, and more. For 1.5 Pro, developers and enterprise customers can also sign up to try a two-million-token context window.

Research

Relentless innovation

Our research team is continually exploring new ideas at the frontier of AI, building innovative products that show consistent progress on a range of benchmarks.

Capability	Benchmark	Description	Gemini 1.5 Flash (May 2024)	Gemini 1.5 Flash (Sep 2024)	Gemini 1.5 Pro (May 2024)	Gemini 1.5 Pro (Sep 2024)
General MMLU-Pro Enhanced version of popular MMLU dataset with questions across multiple subjects with higher difficulty tasks
General	MMLU-Pro	Enhanced version of popular MMLU dataset with questions in 57 subjects (incl. STEM, humanities, and others) with higher difficulty tasks	Gemini 1.5 Flash (May 2024) 59.1%	Gemini 1.5 Flash (Sep 2024) 67.3%	Gemini 1.5 Pro (May 2024) 69.0%	Gemini 1.5 Pro (May 2024) 75.8%
Code Natural2Code Code generation across Python, Java, C++, JS, Go . Held out dataset HumanEval-like, not leaked on the web
Code	Natural2Code	Code generation across Python, Java, C++, JS, Go . Held out dataset HumanEval-like, not leaked on the web	Gemini 1.5 Flash (May 2004) 77.2%	Gemini 1.5 Flash (Sep 2024) 79.8%	Gemini 1.5 Pro (May 2024) 82.6%	Gemini 1.5 Pro (Sep 2024) 85.4%
Math MATH Challenging math problems (incl. algebra, geometry, pre-calculus, and others)
Math	MATH	Challenging math problems (incl. algebra, geometry, pre-calculus, and others)	Gemini 1.5 Flash (May 2004) 54.9%	Gemini 1.5 Flash (Sep 2024) 77.9%	Gemini 1.5 Pro (May 2024) 67.7%	Gemini 1.5 Pro 86.5%
HiddenMath Competition-level math problems, Held out dataset AIME/AMC-like, crafted by experts and not leaked on the web
	HiddenMath	Competition-level math problems, Held out dataset AIME/AMC-like, crafted by experts and not leaked on the web	Gemini 1.5 Flash (May 2004) 20.3%	Gemini 1.5 Flash (Sep 2024) 47.2%	Gemini 1.5 Pro (May 2024) 28.0%	Gemini 1.5 Pro 52.0%
Reasoning GPQA (diamond) Challenging dataset of questions written by domain experts in biology, physics, and chemistry
Reasoning	GPQA (diamond)	Challenging dataset of questions written by domain experts in biology, physics, and chemistry	Gemini 1.5 Flash (May 2024) 41.4%	Gemini 1.5 Flash (Sep 2024) 51.0%	Gemini 1.5 Pro (May 2024) 46.0%	Gemini 1.5 Pro (Sep 2024) 59.1%
Multilingual WMT23 Language translation
Multilingual	WMT23	Language translation	Gemini 1.5 Flash (May 2024) 74.1	Gemini 1.5 Flash (Sep 2024) 73.9	Gemini 1.5 Pro (May 2024) 75.3	Gemini 1.5 Pro (Sep 2024) 75.1
Long Context RULER (at 1M) Diagnostic suite checking long-context ability of the models over a range of tasks
Long Context	RULER (at 1M)	Diagnostic suite checking long-context ability of the models over a range of tasks	Gemini 1.5 Flash (May 2024) 69.6%	Gemini 1.5 Flash (Sep 2024) 82.3%	Gemini 1.5 Pro (May 2024) 40.1%	Gemini 1.5 Pro (Sep 2024) 86.4%
MRCR (1M) Diagnostic long-context understanding evaluation
	MRCR (1M)	Diagnostic long-context understanding evaluation	Gemini 1.5 Flash (May 2024) 70.1%	Gemini 1.5 Flash (Sep 2024) 71.9%	Gemini 1.5 Pro (May 2024) 70.5%	Gemini 1.5 Pro (Sep 2024) 82.6%
Image MMMU Multi-discipline college-level reasoning problems
Image	MMMU	Multi-discipline college-level reasoning problems	Gemini 1.5 Flash (May 2024) 56.1%	Gemini 1.5 Flash (Sep 2024) 62.3%	Gemini 1.5 Pro (May 2024) 62.2%	Gemini 1.5 Pro (Sep 2024) 65.9%
Vibe-Eval (Reka) Visual understanding in chat models with challenging everyday examples. Evaluated with a Gemini Flash model as a rater
	Vibe-Eval (Reka)	Visual understanding in chat models with challenging everyday examples. Evaluated with a Gemini Flash model as a rater	Gemini 1.5 Flash (May 2024) 44.8%	Gemini 1.5 Flash (Sep 2024) 48.9%	Gemini 1.5 Pro (May 2024) 48.9%	Gemini 1.5 Pro (Sep 2024) 53.9%
Image MathVista Mathematical reasoning in visual contexts
	MathVista	Mathematical reasoning in visual contexts	Gemini 1.5 Flash (May 2024) 58.4%	Gemini 1.5 Flash (Sep 2024) 65.8%	Gemini 1.5 Pro (May 2024) 63.9%	Gemini 1.5 Pro (Sep 2024) 68.1%
Audio FLEURS (55 languages) Automatic speech recognition (based on word error rate, lower is better)
Audio	FLEURS (55 languages)	Automatic speech recognition (based on word error rate, lower is better)	Gemini 1.5 Flash (May 2024) 9.8%	Gemini 1.5 Flash (Sep 2024) 9.6%	Gemini 1.5 Pro (May 2024) 6.5%	Gemini 1.5 Pro (May 2024) 6.7%
Video Video-MME Video analysis across multiple domains
Video	Video-MME	Video analysis across multiple domains	Gemini 1.5 Flash (May 2024) 74.7%	Gemini 1.5 Flash (Sep 2024) 76.1%	Gemini 1.5 Pro (May 2024) 77.9%	Gemini 1.5 Pro (May 2024) 78.6%
Safety XSTest Measures how often models refuse to respond to safe/benign prompts. The score represents how frequently models correctly fulfill requests
Safety	XSTest	Measures how often models refuse to respond to safe/benign prompts. The score represents how frequently models correctly fulfill requests	Gemini 1.5 Flash (May 2024) 86.9%	Gemini 1.5 Flash (Sep 2024) 97.0%	Gemini 1.5 Pro (May 2024) 88.4%	Gemini 1.5 Pro (May 2024) 98.8%