gemini-2.5-flash-image

gemini-2.5-flash-image is the image generation and editing model identifier of the Gemini 2.5 Flash series. It is suitable for quickly producing visual sketches, product scene diagrams, social media material variations and image modification solutions in the creative process. It should be used to accurately describe the subject, composition, style, light, size and elements that need to be retained or replaced, and to filter the available results through multiple rounds of generation. It is more suitable for rapid exploration and batch variant production; when it comes to brand logos, real people, copyrighted materials or external content, authorization and manual review are still required.

Context
65536 tokens
Max output
32768 tokens
Input
llm image
Output
llm image

Pricing

Input: $0.3 / 1M tokens

Output: $2.4999999999999 / 1M tokens

Image input: $0.3 / 1M tokens

Image output: $30 / 1M tokens

Endpoints

  • gemini
  • openai

More models

  • gemini-3.1-flash-image

    gemini-3.1-flash-image 是 Gemini 3.1 Flash 系列的图像生成与编辑模型标识,适合需要较快迭代的视觉内容工作流。它可用于将营销创意转成视觉草稿、生成商品背景、改换局部元素、探索配色和制作多版本社交媒体素材。为了提高成片可控性,应先给出清晰的主体与风格约束,再逐步调整构图、光线和细节,而非一次叠加过多要求。涉及可识别人物、品牌资产或公开投放的图片,应加入人工复核和素材权利检查。

  • seedream-4-0-250828

    Seedream 4.0 是字节跳动 Seed 团队开发的图像创作模型,于 2025 年 9 月 9 日正式发布。其核心突破在于采用同一套架构实现文生图与通用编辑能力,融合常识和推理能力,相比前代 Seedream 3.0 和 SeedEdit 3.0,在多模态效果、速度和可用性上均实现显著突破。模型支持文本、图像或任意组合作为输入,覆盖文生图、图生图、单图编辑、多图编辑和图像合成等多种模式,可在 30 余种艺术风格间自由切换,最大分辨率从 2K 提升至 4K,推理速度比 Seedream 3.0 快 10 倍以上。

  • qwen-image-3.0
  • gpt-image-2.5-flare