Gemini Banana 图像生成 (Nano Banana)

Google Gemini 系列同步图像生成 (T2I / I2I / 多图合成 / 多轮编辑, 原生 Interactions API 格式)

POST /v1/images/generations

Auth: {'type': 'bearer', 'description': '在 HTTP header `Authorization: Bearer sk-xxx` 中传入您的 API Key'}

**Gemini Banana (Nano Banana)** 是 Google Gemini 系列的多模态图像生成模型, 支持: - **文生图** (Text-to-Image) - **图像编辑** (Image-to-Image, 传入 1 张参考图 + 文本指令) - **多图合成** (Multi-Image Composition) — `gemini-3.1-flash-image-preview` 支持最多 **14 张参考图** (10 张物体 + 4 张人物) - **多轮编辑** (服务端持有对话历史, 通过 `previous_interaction_id` 链接迭代修改) ## 调用方式 按 **Google Gemini Interactions API** (2026-05 新协议, 字段全 snake_case) 传 body, 同步返回。 客户端继续使用我方 OpenAI 兼容路径 `https://api.router.ai/v1/images/generations`, 不需要直连 Google。 ## 支持的模型 | Model | 说明 | |---|---| | `gemini-3.1-flash-image-preview` | Nano Banana 主力推荐, 4 档分辨率 (512px/1K/2K/4K), 含 thinking 模式, 14 种比例 | | `gemini-3-pro-image-preview` | Nano Banana 专业版, 含 Thinking 推理 + 高保真文本渲染 | | `gemini-2.5-flash-image` | Nano Banana 经典 (GA), 高吞吐低延迟, 1K 单档 | | `gemini-2.5-flash-image-preview` | Nano Banana 经典预览版 (兼容老 SDK 别名) | ## 输出分辨率配置 通过 `response_format.image_size` 指定: | image_size | 实际尺寸 (含 aspect_ratio=1:1) | 适用模型 | |---|---|---| | `"512px"` | 512×512 (注意 **px 后缀**) | 仅 `gemini-3.1-flash-image-preview` | | `"1K"` | 1024×1024 | 全部 (推荐默认) | | `"2K"` | 2048×2048 | gemini-3.1-flash / gemini-3-pro | | `"4K"` | 4096×4096 | gemini-3.1-flash / gemini-3-pro | **`aspect_ratio` 取值** (`gemini-3.1-flash` 支持 14 种): `"1:1"` / `"16:9"` / `"9:16"` / `"3:2"` / `"2:3"` / `"4:3"` / `"3:4"` / `"5:4"` / `"4:5"` + `"1:4"` / `"4:1"` / `"1:8"` / `"8:1"` (仅 `gemini-3.1-flash`) 不传 `response_format` 时走默认 (通常 1K 1:1)。 ## 字段命名 REST API 全部 **snake_case** (`response_format`, `mime_type`, `image_size`, `aspect_ratio`, `previous_interaction_id`, `generation_config`, `system_instruction` 等)。 ## 多轮编辑链 (重要) 服务端默认持有对话历史 (`store: true`)。多轮编辑 1 张图时: 1. 首轮 POST 拿到 `id` (`v1_xxx...`) 2. 第二轮把上次 `id` 放入 `previous_interaction_id` 字段, 重新 POST (无需重传图片, 服务端记得) **数据保留**: 付费层 55 天, 免费层 1 天。 参考: [Google Interactions Image Generation](https://ai.google.dev/gemini-api/docs/interactions/image-generation)

Headers

Request body

Responses

• 200 — 生成成功. Response 含 `id` (多轮用) + `steps[]` (含 model_output 步骤, 内嵌 image content) + `usage` (modality 分)
• 200 — Thinking 模式响应 (steps[] 含 thought 步骤 + model_output, generation_config.thinking_level 设置时返回)
• 200 — 失败状态 (status=failed, steps[] 空; 上游图像安全过滤拦截或 quota 不足)
• 400 — 请求参数错误 (注意 Interactions API 错误信息简洁, 不返 fieldViolations)
• 401 — API Key 无效或缺失
• 503 — 上游服务暂时不可用

Example

# 1. 文生图 (T2I, input 单字符串, 默认 1K 1:1)
curl -X POST https://api.router.ai/v1/images/generations \
  -H "Authorization: Bearer sk-xxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-3.1-flash-image-preview",
    "input": "Create a picture of a nano banana dish in a fancy restaurant with a Gemini theme"
  }'

# 2. T2I + 指定分辨率 2K 16:9 (response_format)
curl -X POST https://api.router.ai/v1/images/generations \
  -H "Authorization: Bearer sk-xxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-3.1-flash-image-preview",
    "input": "a futuristic city at sunset, photorealistic, cinematic",
    "response_format": {
      "type": "image",
      "mime_type": "image/jpeg",
      "aspect_ratio": "16:9",
      "image_size": "2K"
    }
  }'

# 3. I2I 图像编辑 (input 数组, text + image base64)
curl -X POST https://api.router.ai/v1/images/generations \
  -H "Authorization: Bearer sk-xxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-3.1-flash-image-preview",
    "input": [
      {"type": "text", "text": "Create a picture of my cat eating a nano-banana in a fancy restaurant under the Gemini constellation"},
      {"type": "image", "mime_type": "image/jpeg", "data": "<BASE64_IMAGE_DATA>"}
    ]
  }'

# 4. 多图合成 (Multi-Image Composition, 多张参考图)
curl -X POST https://api.router.ai/v1/images/generations \
  -H "Authorization: Bearer sk-xxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-3.1-flash-image-preview",
    "input": [
      {"type": "text", "text": "An office group photo of these people, they are making funny faces."},
      {"type": "image", "mime_type": "image/png", "data": "<BASE64_DATA_IMG_1>"},
      {"type": "image", "mime_type": "image/png", "data": "<BASE64_DATA_IMG_2>"},
      {"type": "image", "mime_type": "image/png", "data": "<BASE64_DATA_IMG_3>"}
    ],
    "response_format": {
      "type": "image",
      "mime_type": "image/jpeg",
      "aspect_ratio": "5:4",
      "image_size": "2K"
    }
  }'

# 5. 多轮编辑链 (上次 id 放入 previous_interaction_id, 无需重传图片)
curl -X POST https://api.router.ai/v1/images/generations \
  -H "Authorization: Bearer sk-xxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-3.1-flash-image-preview",
    "input": "Update this infographic to be in Spanish. Do not change any other elements of the image.",
    "previous_interaction_id": "v1_ChdEdGNPYXY3bkhNV00xZThQNjlyaXdBMBIXRHRjT2F2N25ITVdNMWU4UDY5cml3QTA",
    "response_format": {
      "type": "image",
      "mime_type": "image/jpeg",
      "aspect_ratio": "16:9",
      "image_size": "2K"
    }
  }'

# 6. Thinking 模式 (gemini-3.1-flash / gemini-3-pro 支持; minimal 默认 / high 推理强度)
curl -X POST https://api.router.ai/v1/images/generations \
  -H "Authorization: Bearer sk-xxx" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-3-pro-image-preview",
    "input": "draw a step-by-step guide on how to bake a banana bread",
    "generation_config": {
      "thinking_level": "high"
    },
    "response_format": {
      "type": "image",
      "image_size": "1K"
    }
  }'

API reference