Gemini Models  |  Gemini API  |  Google AI for Developers

Last updated: 3/21/2026

Try the new Built-in tools and Function calling combination feature, and Grounding with Google Maps, now supported for Gemini 3 models.

Was this helpful?

Send feedback

Models

Gemini 3

auto_awesome Gemini 3.1 Pro\ \ Advanced intelligence, complex problem-solving skills, and powerful agentic and vibe coding capabilities.\ \ \ New Preview spark Gemini 3 Flash\ \ Frontier-class performance rivaling larger models at a fraction of the cost.\ \ \ Preview bolt Gemini 3.1 Flash-Lite\ \ Frontier-class performance rivaling larger models at a fraction of the cost.\ \ \ New Preview 🍌🍌 Nano Banana 2\ \ Powerful, high-efficiency image generation and editing, optimized for speed and high-volume use cases.\ \ \ Preview 🍌 Nano Banana Pro\ \ State-of-the-art image generation and editing models for highly contextual native image creation.\ \ \ Preview

Gemini 2.5 Flash

bolt

Gemini 2.5 Flash

Our best price-performance model for low-latency, high-volume tasks that require reasoning.

image_search

Nano Banana

State-of-the-art native image generation and editing designed for fast, creative workflows.

settings_voice

Gemini 2.5 Flash Live Preview

Optimized for real-time conversational agents with sub-second native audio streaming.

record_voice_over

Gemini 2.5 Flash TTS Preview

Controllable text-to-speech audio generation with fine control over style and pacing.

Gemini 2.5 Flash-Lite

speed

Gemini 2.5 Flash-Lite

The fastest and most budget-friendly multimodal model in the 2.5 family.

Gemini 2.5 Pro

psychology

Gemini 2.5 Pro

Our most advanced model for complex tasks, featuring deep reasoning and coding capabilities.

record_voice_over

Gemini 2.5 Pro TTS Preview

High-fidelity speech synthesis optimized for quality in structured workflows like podcasts and audiobooks.

Audio models

This section contains all audio models, including ones that may already be listed in other sections

settings_voice

Gemini 2.5 Flash Live Preview

Our flagship Live API model for low-latency, bidirectional voice and video agents with native audio reasoning.

bolt

Gemini 2.5 Flash TTS Preview

Fast and controllable text-to-speech for low-latency, cost-efficient applications and real-time assistants.

record_voice_over

Gemini 2.5 Pro TTS Preview

High-fidelity speech synthesis optimized for quality in structured workflows like podcasts and audiobooks.

music_note

Lyria Experimental

High-fidelity music generation model providing granular creative control over instruments, BPM, and complex compositions.

Generative media models

This section contains all generative media models, including ones that may already be listed in other sections

blur_on

Nano Banana 2 Preview

High-efficiency production-scale visual creation, combining the intelligence of the Gemini 3 series with lightning-fast generation speeds.

movie_filter

Veo 3.1 Preview

State-of-the-art cinematic video generation with advanced creative controls and natively synchronized audio.

image_edit

Nano Banana Pro Preview

A professional design engine with a reasoning core for studio-quality 4K visuals, complex layouts, and precise text rendering.

music_note

Lyria Experimental

High-fidelity music generation model providing granular creative control over instruments, BPM, and complex compositions.

image_search

Nano Banana

State-of-the-art native image generation and editing designed for fast, creative workflows.

palette

Imagen 4

Text-to-image model yet, featuring fast and ultra-fast generation and exceptional clarity up to 2K resolution.

Tool and agent models

desktop_windows

Computer Use Preview

A specialized model that can "see" a digital screen and perform UI actions like clicking, typing, and navigating to automate complex browser tasks.

travel_explore

Gemini Deep Research Preview

An agentic model that autonomously plans and executes multi-step research across hundreds of sources to produce cited, interactive reports.

Specialized task models

hub

Gemini Embedding 2 Preview

Our first multimodal embedding model, mapping text, images, video, audio, and PDFs into a unified embedding space for advanced semantic search and RAG systems.

hub

Gemini Embedding

High-dimensional vector representations for advanced semantic search, text classification, and RAG systems.

precision_manufacturing

Gemini Robotics Preview

Advanced embodied reasoning model that understands physical spaces and plans multi-step tasks for robotic agents.

Previous models

history

Gemini 2.0 Flash Deprecated

Our second generation workhorse model, with next-gen features and improved capabilities, including superior speed, native tool use, and a 1M token context window.

restore

Gemini 2.0 Flash-Lite Deprecated

Our fastest second generation model, optimized for cost efficiency and low latency.

restore

Gemini 3 Pro Preview Shut down

Our state-of-the-art reasoning model, with advanced multimodal understanding.

Model version name patterns

Gemini models are available in either stable, preview, latest, or experimental versions.

Stable

Points to a specific stable model. Stable models usually don't change. Most production apps should use a specific stable model.

For example: gemini-2.5-flash.

Preview

Points to a preview model which may be used for production. Preview models will typically have billing enabled, might come with more restrictive rate limits and will be deprecated with at least 2 weeks notice.

For example: gemini-2.5-flash-preview-09-2025.

Latest

Points to the latest release for a specific model variation. This can be a stable, preview or experimental release. This alias will get hot-swapped with every new release of a specific model variation. A 2-week notice will be provided through email before the version behind latest is changed.

For example: gemini-flash-latest.

Experimental

Points to an experimental model which will typically be not be suitable for production use and come with more restrictive rate limits. We release experimental models to gather feedback and get our latest updates into the hands of developers quickly.

Experimental models are not stable and availability of model endpoints is subject to change.

Model deprecations

For information about model deprecations, visit the Gemini deprecations page.

Was this helpful?

Send feedback

Related Articles