模型广场

Gemini 3.7 Flash is Google's multimodal workhorse model for fast agentic workflows, coding, and complex multi-step reasoning, improving on 3.6 Flash in agentic benchmarks (GPQA Diamond ~94%, TAU-Bench ~80%). All input modalities (text/image/video/audio) unified pricing; batch tier at 50% discount. Thinking model with configurable levels. 1M context. Released August 13, 2026.

Gemini 3.5 Flash-Lite is Google's high-throughput, low-latency multimodal model with upgraded agentic capabilities, suited for subagents executing focused tasks within complex multi-agent workflows, agentic search, and document processing. All input modalities (text/image/video/audio) unified at $0.30/1M; batch tier at 50% discount. Successor to Gemini 3.1 Flash Lite. Released July 21, 2026.

Gemini 3.6 Flash is Google's token-efficient workhorse model: 17% fewer output tokens than 3.5 Flash, fewer reasoning steps and tool calls in multi-step agentic workflows, and higher-precision code edits with reduced execution loops. All input modalities (text/image/video/audio) unified at $1.50/1M; batch tier at 50% discount. Thinking model with configurable levels. Released July 21, 2026.

Gemini 3.1 Flash Lite Image is Google's most cost-efficient image generation and editing model, supporting text-to-image, image editing, and multi-image composition. Outputs generated at 1K resolution across 14 aspect ratios, at the lowest price point in the Nano Banana family.

Nano Banana Pro is Google's most advanced image-generation and editing model, built on Gemini 3 Pro. GA release of gemini-3-pro-image-preview. It generates context-rich graphics from infographics and diagrams to cinematic composites, with 2K/4K output, multi-image blending, identity preservation, and localized edits.

Gemini 3.1 Flash Image, a.k.a. "Nano Banana 2," is Google's state of the art image generation and editing model, delivering Pro-level visual quality at Flash speed. GA release of gemini-3.1-flash-image-preview, combining advanced contextual understanding with fast, cost-efficient inference.

Gemini 3.5 Flash is Google's efficient multimodal model delivering near-Pro level coding and reasoning at Flash-tier cost and speed. Optimized for coding tasks and parallel agent execution with configurable thinking levels. Released May 20, 2026.

Gemini 3.1 Flash Lite (GA) is Google's high-efficiency multimodal model optimized for low-latency, high-volume workloads. GA version of the preview model. Supports full thinking levels (minimal, low, medium, high) for cost/performance trade-offs. Priced at half the cost of Gemini 3 Flash. Released May 7, 2026.

Gemini 3.1 Pro is the next generation in the Gemini series of models, a suite of highly-capable, natively multimodal, reasoning models. Gemini 3 Pro is now Google’s most advanced model for complex tasks, and can comprehend vast datasets, challenging problems from different information sources, including text, audio, images, video, and entire code repositories

Gemini 3 Flash Preview is a high speed, high value thinking model designed for agentic workflows, multi turn chat, and coding assistance. It delivers near Pro level reasoning and tool use performance with substantially lower latency than larger Gemini variants, making it well suited for interactive development, long running agent loops, and collaborative coding tasks. Compared to Gemini 2.5 Flash, it provides broad quality improvements across reasoning, multimodal understanding, and reliability.

Gemini 2.5 Flash Image, a.k.a. "Nano Banana," is now generally available. It is a state of the art image generation model with contextual understanding. It is capable of image generation, edits, and multi-turn conversations.

Gemini 2.5 Flash-Lite is a lightweight reasoning model in the Gemini 2.5 family, optimized for ultra-low latency and cost efficiency. It offers improved throughput, faster token generation, and better performance across common benchmarks compared to earlier Flash models. By default, [thinking] (i.e. multi-pass reasoning) is disabled to prioritize speed, but developers can enable it via the Reasoning API parameter to selectively trade off cost for intelligence.

Fast and cost-effective Gemini 2.5 with 1M context. Features toggleable reasoning capabilities and full multimodal support at significantly lower cost than Pro.

Google's flagship Gemini model with 1M context, thinking/reasoning capabilities, and full multimodal support (text, image, video, audio). Excels at complex analysis and creative tasks.