gemini-3-pro-image AI 图像 API
gemini-3-pro-image (Nano Banana Pro): 通过 Flatkey 生成图片、参考图编辑并迭代视觉设计。下方示例使用 Gemini contents/parts 契约,不能直接替换成 Images API 请求体。
{
"model": "gemini-3-pro-image",
"contents": [
{
"role": "user",
"parts": [
{
"text": "Scene: A sepia archival photograph shows four workers weaving cloth on wooden looms inside a rustic workshop, with yarn baskets and bright hillside scenery through the windows. Restore the paper texture, faces without sharpening them into modern portraits, loom construction, period clothing, and gentle film wear. No modern objects, readable signage, logos, captions, or watermark.\n\nComposition and restoration: Deliver one 16:9 archival-style still. Preserve the original framing, period-specific construction and restrained tonal range; retain monochrome, sepia or faded color as described. Recover modest detail in faces, clothing and architecture without synthetic sharpening. Keep fine grain and authentic aging consistent across the frame; repair damage without replacing historical features.\n\nAvoid: warped perspective, malformed anatomy, merged objects, inconsistent shadows, excessive sharpening and watermarks. Follow the scene-specific text and branding restrictions above."
}
]
}
],
"generationConfig": {
"responseModalities": [
"TEXT",
"IMAGE"
],
"imageConfig": {
"aspectRatio": "16:9",
"imageSize": "1K"
}
}
}API 性能
gemini-3-pro-image API 可靠性与可用性
当生产流量达到足够规模后,这里会显示 gemini-3-pro-image 的 Flatkey 实时请求遥测数据。
API 活动
gemini-3-pro-image API 使用量与请求活动
查看最近统计窗口内 gemini-3-pro-image 的请求量和成功推理活动。
gemini-3-pro-image 价格
gemini-3-pro-image API 价格与计费
价格卡展示 Flatkey 当前输入、输出 token 等费率。Token 单价不是每张图片的固定报价:图片设置、参考输入与生成内容都会影响用量。Google 直连价格和免费额度不自动适用于此接口。
充值余额
图像、视频、音频和文本模型共用同一个 Flatkey 余额和 API Key。
- 模型类型
- 文生图
- API 接入
- /v1beta/models/gemini-3-pro-image:generateContent
- 计费依据
- 1M tokens
- 输出数量
- 1
gemini-3-pro-image AI 图像生成器与 API: 核心能力
Google 的 gemini-3-pro-image 路由支持 文本 · 图像。目录分类:Marketing。
文生图
描述主体、构图、光线和需要生成的文字。尺寸等输出控制应放在请求字段中,不要只写进提示词。
参考图片
提供参考图片,并明确需要保留和修改的部分。参考素材数量以本页参数矩阵为准。
输出数量
只选择当前模型 Playground 提供的比例和图片尺寸。Flash、Flash Lite 和 Pro 的输出限制不能混用。
API 接入
保留下方准确的模型 ID 与 contents/parts 请求结构。接入批量图片任务前,先用小请求验证。
对比
gemini-3-pro-image API 对比:能力与访问方式
在调整集成前,将 gemini-3-pro-image 的目录事实与上一代基线进行对比。
| 能力 | 上一代 | gemini-3-pro-image |
|---|---|---|
| 供应商 | 此目录快照未验证 | |
| 模态 | 此目录快照未验证 | 文本 · 图像 |
| 上下文 | 此目录快照未验证 | 目录未列出上下文窗口 |
| 端点 | 此目录快照未验证 | /v1beta/models/gemini-3-pro-image:generateContent |
| 计费 | 此目录快照未验证 | $2/$12 per 1M tokens |
接入前先选对请求方式
核对支持的工作流、当前配置的实时费用,以及 Flatkey 实际使用的参数契约。
当目录提供固定单价时,根据当前 Playground 配置和实时目录价格计算。
- 当前配置
- 1×
- 实时单价
- $2
- 本次预估
- 按实际用量计费
最终费用以实际提交的请求和账户分组为准。Token 或分层计费无法简化成一个固定预估值。
工作流与 Model ID
只使用标记为可用的工作流。Flatkey 对外保持一个模型 ID,由请求体选择输入方式。
文生图
通过文本提示词和输出参数生成新图片。
- 模型 ID
gemini-3-pro-image- 接口
/v1beta/models/gemini-3-pro-image:generateContent
参考图生成与编辑
用参考图片完成编辑、变体或主体与构图引导。
- 模型 ID
gemini-3-pro-image- 接口
/v1beta/models/gemini-3-pro-image:generateContent
参数兼容矩阵
以下取值与 Playground 和请求预览使用同一份模型配置。
aspect_ratio画面比例1:1 · 2:3 · 3:2 · 3:4 · 4:3 · 4:5 · 5:4 · 9:16 · 16:9 · 21:9
默认值: 1:1image_size尺寸1K · 2K · 4K
默认值: 1Kimage参考素材上限4 个文件
接口: /v1beta/models/gemini-3-pro-image:generateContent从首次测试到生产接入
Playground 与 API 使用同一套请求契约,并明确处理参数校验、余额、任务状态和产物获取。
请求生命周期
- 1
校验模型、提示词、参考素材和输出设置。
- 2
使用已文档化字段提交图片请求。
- 3
读取路由返回的生成结果或异步任务结果。
- 4
在自己的工作流中保存生成文件和请求元数据。
生产检查项
提交前拦截不支持的尺寸、时长、比例、格式或参考素材数量。
分别处理无效密钥、余额不足、限流和账户分组价格。
明确指出被拒绝的提示词或素材,让用户只替换失败的输入。
保留任务 ID,限制重试次数,且不能把超时当成生成成功。
提示词库
gemini-3-pro-image 图像生成提示词示例
先使用一个面向图像生成的 gemini-3-pro-image 提示词,再调整 Playground 中显示的请求字段。
使用 Image 2 生成
Scene: A 16:9 underwater game loadout screen on a sea floor. An armored diver stands on a circular pedestal holding a glowing cyan sword, with submerged ruins behind and a weapon-swap panel on the right. Lock the diver silhouette, sword, aquatic lighting, pedestal, and interface hierarchy for an exact equipment-switch frame. No readable words, logos, invented stats, or watermark. Composition and rendering: Deliver one 16:9 game-interface keyframe. Separate the character, held equipment and HUD into clear depth layers; keep slot spacing and icon scale consistent, with the selected item visually dominant. Preserve the scene-specific game art style. Let material edges catch the existing light without letting interface glow obscure the silhouette. Keep all equipment fully inside the frame. Avoid: warped perspective, malformed anatomy, merged objects, inconsistent shadows, excessive sharpening and watermarks. Follow the scene-specific text and branding restrictions above.
使用 Image 2 生成
Scene: A televised basketball frame shows a red-uniform player taking a jump shot while a white-uniform defender contests it in an indoor arena. Include the hoop, overhead lights, crowd blur, and abstract red-and-gray lower-third panels without readable text. Keep the broadcast angle, ball arc, uniforms, and player spacing coherent; no real teams, athletes, logos, or watermark. Composition and rendering: Deliver one 16:9 broadcast still at the decisive instant. Use coherent perspective for the playing surface and equipment, readable separation between competitors, and a softer crowd behind the action. Keep the principal subject sharp, with directional blur limited to fast extremities and spray. Preserve the specified broadcast angle and reserve overlay areas clear of the action. Avoid: warped perspective, malformed anatomy, merged objects, inconsistent shadows, excessive sharpening and watermarks. Follow the scene-specific text and branding restrictions above.
使用 Image 2 生成
Scene: A clear glass dropper bottle filled with pale gold serum stands on a white stone block in a warm cream room. Soft sunlight and large leaf shadows fall across the wall and the liquid, glass, metal collar, and surface catch subtle reflections. Preserve the bottle proportions and generous blank space; no readable branding, extra products, hands, or watermark. Composition and rendering: Deliver one 16:9 commercial product still. Keep the complete product silhouette inside generous crop-safe margins. Resolve edges, seams and material transitions precisely; shape highlights to describe volume without clipping bright surfaces or crushing dark details. Align reflections with the object and light source. Retain the stated camera angle and existing negative space; add no decorative props. Avoid: warped perspective, malformed anatomy, merged objects, inconsistent shadows, excessive sharpening and watermarks. Follow the scene-specific text and branding restrictions above.
使用 Image 2 生成
Scene: At a rain-wet historic train platform, a young traveler in a dark cap and coat holds a brown suitcase while a steam train waits under a long canopy. Warm lamps recede into the blue night and reflect on the platform. Hold the traveler profile, suitcase, train direction, and wide composition for a continuous storyboard. No readable signs, logos, or watermark. Composition and rendering: Deliver one 16:9 storyboard keyframe, not a collage or an action sequence. Separate foreground, character plane and background through the existing light and atmospheric depth. Keep the stated pose and eyelines readable, and retain environmental landmarks for continuity. Match the described photographic, illustrated or puppet medium; give fabric and surfaces texture appropriate to that medium. Avoid: warped perspective, malformed anatomy, merged objects, inconsistent shadows, excessive sharpening and watermarks. Follow the scene-specific text and branding restrictions above.
使用 Image 2 生成
Scene: On a small theater stage, a startled magician in black pulls a long rainbow cloth from a tall black trunk wrapped in more cloth. The audience is dim behind him, with red curtains and colored spotlights. Freeze the harmless reveal at the moment the fabric stretches between hand and trunk; preserve object weight, pose, and stage spacing. No readable words, logos, injury, or watermark. Composition and rendering: Deliver one 16:9 physical-comedy still at the described instant. Make the cause of the gag readable through body balance, grip and object spacing. Separate the reaction from the airborne props; retain believable gravity and contact points. Keep the face and important props sharp, with only restrained directional blur on fast edges. Preserve the stated lighting and room or stage layout. Avoid: warped perspective, malformed anatomy, merged objects, inconsistent shadows, excessive sharpening and watermarks. Follow the scene-specific text and branding restrictions above.
使用 Image 2 生成
Scene: A sepia archival photograph shows four workers weaving cloth on wooden looms inside a rustic workshop, with yarn baskets and bright hillside scenery through the windows. Restore the paper texture, faces without sharpening them into modern portraits, loom construction, period clothing, and gentle film wear. No modern objects, readable signage, logos, captions, or watermark. Composition and restoration: Deliver one 16:9 archival-style still. Preserve the original framing, period-specific construction and restrained tonal range; retain monochrome, sepia or faded color as described. Recover modest detail in faces, clothing and architecture without synthetic sharpening. Keep fine grain and authentic aging consistent across the frame; repair damage without replacing historical features. Avoid: warped perspective, malformed anatomy, merged objects, inconsistent shadows, excessive sharpening and watermarks. Follow the scene-specific text and branding restrictions above.
为什么使用 gemini-3-pro-image API
为什么通过 Flatkey 使用 gemini-3-pro-image?
通过一个网关管理 gemini-3-pro-image、账户控制以及模型目录中的其他模型。
一个 API 使用 gemini-3-pro-image
明确保留 gemini-3-pro-image 模型 ID 和端点,同时与其他工作负载共用 Flatkey 密钥。
实时目录价格
发送请求前查看当前 gemini-3-pro-image 计费维度,并在账户中确认最终预估。
实用的模型交接
在公开 Playground 中测试 gemini-3-pro-image 提示词,再将相同设置带入经过身份验证的集成。
用量与路由控制
集中管理 gemini-3-pro-image 的密钥、配额和路由,无需改变应用面向供应商的工作流。
API 接入
gemini-3-pro-image API
四种接入方式共用同一个 Key 和模型目录,选择一种查看可运行示例。
curl https://router.flatkey.ai/v1beta/models/gemini-3-pro-image:generateContent \ -H "Content-Type: application/json" \ -H "Authorization: Bearer sk-***" \ -d '{"model":"gemini-3-pro-image","contents":[{"role":"user","parts":[{"text":"使用 gemini-3-pro-image 制作高质量产品图:构图干净、光线精准、主体突出、细节真实。"}]}],"generationConfig":{"responseModalities":["TEXT","IMAGE"],"imageConfig":{"aspectRatio":"1:1","imageSize":"1K"}}}'
常见问题
gemini-3-pro-image API
价格、功能与使用常见问题
关于 gemini-3-pro-image 价格、能力、端点访问和目录限制的答案。





