Gemini Models: All Google Model Overview and Key Features

Google's Gemini model family has grown significantly since its launch, spanning multiple generations and capability tiers. From high-throughput Flash-Lite models to frontier reasoning with Pro, there is a Gemini model for nearly every use case and budget. This guide lists every major Google Gemini model in order from the newest to the oldest, with a concise overview of each model's specs, key features, and intended use — plus links to try the ones available on Chat Smith.

Released July 21, 2026, Gemini 3.6 Flash is Google's latest mid-tier workhorse. It matches Gemini 3.5 Flash on raw intelligence (Artificial Analysis Index: 50) but delivers roughly 17% fewer output tokens for the same tasks and cuts output pricing from $9.00 to $7.50 per million tokens. Knowledge cutoff advances to March 2026. Benchmarks show meaningful improvements in agentic coding (DeepSWE: 49% vs. 37%), computer use (OSWorld: 83% vs. 78.4%), and ML research (MLE-Bench: 63.9% vs. 49.7%). Context window: 1M tokens. Speed: ~275 tok/s. Best for: agentic coding, document analysis, knowledge work. → Read Gemini 3.6 Flash full review

Also launched July 21, 2026, Gemini 3.5 Flash-Lite is the fastest and most cost-effective model in the 3.5 class. It runs at 350 output tokens per second and is priced at $0.30/1M input and $2.50/1M output tokens. Despite the lower price, it significantly outperforms 3.1 Flash-Lite on agentic and coding benchmarks (SWE-Bench Pro: 54.2% vs. 31%). Configurable thinking levels (minimal to high) let developers balance speed against quality. Best for: high-throughput pipelines, agentic search, document processing. → Read Gemini 3.5 Flash-Lite full review

Launched at Google I/O 2026, Gemini 3.5 Flash is a frontier-class agent model built for complex long-horizon tasks. It rivals large flagship models on coding and agentic benchmarks while maintaining the speed expected from the Flash series. Multimodal input (text, image, audio, video), 1M token context, and strong instruction following make it a versatile production choice. Pricing: $1.50/1M input, $9.00/1M output. → Read Gemini 3.5 Flash review

Gemini 3.1 Flash-Lite is Google's most cost-efficient model in the 3-series, optimized for high-volume, latency-sensitive tasks. It adds flexible thinking levels (minimal to high) for balancing speed and quality, improved audio input, and matches Gemini 2.5 Flash on key capabilities. Pricing: $0.25/1M input, $1.50/1M output. Speed: ~363 tok/s. Context: 1M tokens. Best for: high-volume classification, chatbots, instruction-heavy workflows. → Read Gemini 3.1 Flash-Lite review

Gemini 3.1 Pro is the flagship model of the 3.1 generation, positioned for the most complex reasoning, coding, and long-context tasks. It scores 46 on the Artificial Analysis Intelligence Index and targets enterprise use cases requiring high accuracy at scale. Pricing: ~$2.00/1M input, ~$12.00/1M output. Speed: ~145 tok/s. Context: 1M+ tokens. Best for: deep reasoning, complex multi-step workflows, enterprise pipelines. → Read Gemini 3.1 Pro review

Gemini 3 Flash was designed as a fast, efficient model combining speed with strong multimodal capabilities. It handles text, image, audio, and video inputs across a 1M token context window, and was widely adopted for production applications before the 3.5 generation arrived. Best for: general-purpose chat, summarization, classification, and rapid prototyping. → Read Gemini 3 Flash review

Gemini 3 Pro offers the highest capability in the Gemini 3 generation, targeting frontier reasoning and complex task execution. It expands on the Flash's multimodal strengths with greater depth in analysis, instruction following, and long-context understanding. Best for: deep research, complex document workflows, and high-stakes reasoning tasks. → Read Gemini 3 Pro review

Nano Banana is Google's native image generation and editing model, officially part of the Gemini image model family. It supports multi-image fusion, character consistency, style control, and iterative editing within a conversational workflow. Two variants are available: Nano Banana (standard) and Nano Banana Pro (higher quality output), both available on Chat Smith. Best for: AI image generation, photo editing, brand visuals, character creation.

Gemini 2.5 Pro is Google's first hybrid reasoning model with a 1M token context window and configurable thinking budgets. It delivers strong performance on complex coding, analysis, and multi-step reasoning tasks, and was widely regarded as one of the top models in its class at launch. Pricing: $1.25/1M input, $10.00/1M output. Best for: complex reasoning, coding, document analysis. → Read Gemini 2.5 Pro review

Gemini 2.5 Flash is a fast, capable production model balancing cost and performance for high-volume applications. It supports a 1M token context window, multimodal inputs, and a generous free tier via the Gemini API. Pricing: $0.30/1M input, $2.50/1M output. Speed: ~249 tok/s. Best for: chatbots, content classification, translation, and rapid prototyping. → Read Gemini 2.5 Flash review

Gemini 2.5 Flash-Lite is Google's budget champion — the cheapest model with a 1M token context window on the market at $0.10/1M input and $0.40/1M output. At ~366 tok/s, it is also among the fastest models available. Best for: cost-sensitive, high-volume workloads where speed and affordability matter more than deep reasoning. → Read Gemini 2.5 Flash-Lite review

Gemini 2.0 Flash was Google's previous-generation workhorse, known for fast, multimodal responses across text, image, audio, and video. It was deprecated on June 1, 2026 and is no longer available via the Gemini API. Teams that were using it are now advised to migrate to Gemini 3 Flash or 3.1 Flash-Lite. Best for: reference only — migrate to a newer model. → Read Gemini 2.0 Flash review

Released in May 2024, Gemini 1.5 Flash was a landmark model at launch — the first Flash-class model to offer a 1M token context window at developer-friendly pricing. Built via distillation from Gemini 1.5 Pro, it supports text, image, audio, and video inputs with ~190 tok/s throughput. Pricing: $0.075–$0.15/1M input, $0.30–$0.60/1M output. While superseded by newer generations, it remains a useful historical reference for understanding the Flash model lineage. → Read Gemini 1.5 Flash review

Google's Gemini lineup now spans three major generations and multiple capability tiers, from budget Flash-Lite models to frontier Pro reasoning. For developers and teams looking to work with these models without managing API keys, Chat Smith gives you access to the full range of available Gemini models alongside GPT, Claude, Grok, and DeepSeek — all from a single interface. Browse the full model catalog on Chat Smith to find the right Gemini model for your workflow.

1. What is the newest Gemini model in 2026?

As of July 2026, the newest Gemini models are Gemini 3.6 Flash and Gemini 3.5 Flash-Lite, both released on July 21, 2026. Gemini 3.6 Flash is the primary workhorse with improved coding and knowledge work performance, while Gemini 3.5 Flash-Lite is the fastest and most cost-effective model at 350 tokens per second.

2. Which Gemini models are available on Chat Smith?

Chat Smith currently offers Gemini 3.5 Flash, Gemini 3.1 Flash-Lite, Gemini 3 Flash, Gemini 2.5 Pro, Gemini 2.5 Flash, Gemini 2.5 Flash-Lite, Nano Banana, and Nano Banana Pro. You can browse the full list on the Chat Smith model page.

3. What is the difference between Gemini Flash and Gemini Pro models?

Flash models prioritize speed and cost efficiency, making them ideal for high-volume, latency-sensitive production applications. Pro models offer deeper reasoning, stronger performance on complex tasks, and higher accuracy at the cost of slower speeds and higher pricing. Flash-Lite models extend this further toward maximum throughput at minimum cost.