top of page
Search

All Gemini models available in 2025: complete list for web, app, API, and Vertex AI

ree

Gemini now includes over ten model variants across web, mobile, and API access points.

As of August 2025, Google Gemini has expanded into a complex ecosystem of specialized models, accessible via the Gemini web/app interface, Google AI Studio, and Google Cloud’s Vertex AI. The current offerings reflect Google's multi-layered strategy: lightweight models for speed, advanced models for reasoning, and dedicated APIs for real-time voice and multimodal interaction.

Here we share a complete breakdown of every Gemini model currently available, sorted by use case and deployment method.



Gemini web and mobile apps include both Pro and Flash families.

The public-facing Gemini experience (via gemini.google.com or the Android/iOS apps) offers both free and premium access tiers, depending on subscription type (Google One AI Premium, Workspace Pro/Ultra, etc.). These are the active model families:


● Gemini 2.5 Flash (experimental rollout)

Now in progressive rollout on the app, this model emphasizes fast, grounded answers with conversational refinement. It replaces Gemini 1.5 Flash and is optimized for general queries and search tasks. While it lacks reasoning depth, it is ideal for low-latency user interaction.


● Gemini 2.5 Pro (experimental rollout)

Available to AI Premium or select Workspace users, this version includes thinking mode and performs better in multi-step reasoning, summarization, and creative generation. It is gradually replacing older Pro variants in user-facing applications.


● Gemini 2.0 Flash

The fallback or standard model still deployed in many regions. It remains active in the mobile app for regular queries and light reasoning. Some systems retain it for stability and compatibility.


● Gemini Live (voice/video model)

This refers to a special class of models optimized for real-time interaction via voice and camera. Available in over 150 countries and 45+ languages, Gemini Live offers spoken conversations, facial tracking, and visual input. The actual backend varies by session but is part of the 2.5 Flash Live family.



Google AI Studio and Gemini API support multiple versions with fine-grained control.

Google’s developer console (AI Studio and Gemini API) provides direct access to model endpoints, each with a specific model code and functionality profile. These are the current models available for production and preview use.

Model Name

Model Code

Status

Gemini 2.5 Pro

gemini-2.5-pro

Stable

Gemini 2.5 Flash

gemini-2.5-flash

Stable

Gemini 2.5 Flash-Lite

gemini-2.5-flash-lite

Stable (GA)

Gemini 2.5 Flash Live (Preview)

gemini-live-2.5-flash-preview

Preview

Gemini 2.5 Native Audio Dialog

gemini-2.5-flash-preview-native-audio-dialog

Experimental

Gemini 2.5 TTS (text-to-speech)

gemini-2.5-flash-preview-tts, gemini-2.5-pro-preview-tts

Preview

Gemini 2.0 Flash

gemini-2.0-flash

Stable

Gemini 2.0 Flash-Lite

gemini-2.0-flash-lite

Stable

Gemini 2.0 Flash Live

gemini-2.0-flash-live-001

General Access

Gemini 2.0 Flash (image generation)

gemini-2.0-flash-preview-image-generation

Preview

Gemini 1.5 Pro / Flash / Flash-8B

gemini-1.5-pro, gemini-1.5-flash, gemini-1.5-flash-8b

Deprecated since Apr 29, 2025

The model codes listed above are used when deploying tasks via the Gemini API or Google’s SDKs. 2.5 Flash-Lite, in particular, has become a popular choice for mobile apps and embedded systems due to its cost-efficiency and low-latency behavior.



Vertex AI offers 2.5 models with full production support.

On Google Cloud's Vertex AI, the main Gemini variants are available as fully integrated managed models with production-grade support. These include:


● Gemini 2.5 Pro (GA)

The most powerful and complete Gemini model currently available. It supports contextual thinking, long-form generation, and stable performance for enterprise scenarios.


● Gemini 2.5 Flash and Flash-Lite (GA)

Both variants are now generally available, with Flash-Lite specifically promoted to GA status in late July 2025. These are the preferred options for building scalable AI services, especially when performance and pricing need to be balanced.


Deprecated models like Gemini 1.5 Pro and 1.5 Flash were fully retired for new projects as of April 29, 2025, although existing users may retain access for legacy integrations.


Choosing between Gemini models depends on interface, latency, and reasoning needs.

Gemini's model ecosystem is now functionally segmented by access tier and usage channel. Here is a general guide for choosing the right variant:

  • For fast answers or search-style conversations → use Gemini 2.5 Flash or 2.5 Flash-Lite

  • For multi-step thinking, summaries, or code tasks → use Gemini 2.5 Pro

  • For real-time voice interaction → use Gemini Live via mobile or native audio API

  • For image editing/generation → use 2.0 Flash image preview model

  • For cloud integration with guaranteed support → deploy on Vertex AI (2.5 GA models)


All Gemini 2.5 models include improved multilingual support, better safety filters, and flexible deployment options across mobile, web, and cloud services. For developers, the Gemini API remains the most granular way to control model behavior, latency, and cost.



____________

FOLLOW US FOR MORE.


DATA STUDIOS


bottom of page