Image Generation

AnyInt publishes Gemini-native, DashScope, and Transtreams Kling image-generation paths. They overlap in outcome, but the request bodies and strengths are different.

Published routes

Route	Notes
`/gemini/v1beta/models/gemini-3.1-flash-image-preview:generateContent`	Requires `responseModalities` to include `TEXT` and `IMAGE`
`/dashscope/v1/services/aigc/multimodal-generation/generation`	Prompt-driven image generation with parameters such as `size`, `negative_prompt`, and `watermark`
`/transtreams/kling/v1/images/generations`	Kling v2.1 image generation task route
`/transtreams/kling/v1/images/omni-image`	Kling V3 Omni image generation task route

Use Gemini when

you want a Gemini-native request shape
you want mixed text-and-image output
you are already building around Gemini contents[].parts[]

Use DashScope when

you want parameters such as size, negative_prompt, or watermark
you are already integrating other DashScope media flows
you want a more explicitly media-oriented payload instead of a Gemini conversation format

Use Kling when

you want kling-v2-1, displayed as Kling: Image v2.1
you want the Omni image route for kling-v3-omni
you are building an async image workflow and can poll task results

Gemini example

curl https://gateway.api.anyint.ai/gemini/v1beta/models/gemini-3.1-flash-image-preview:generateContent \
  -H "Authorization: Bearer $ANYINT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [
      {
        "role": "user",
        "parts": [
          {"text": "Generate a launch poster for an AI music product, cinematic and high-contrast."}
        ]
      }
    ],
    "generationConfig": {
      "responseModalities": ["TEXT", "IMAGE"]
    }
  }'

DashScope example

curl https://gateway.api.anyint.ai/dashscope/v1/services/aigc/multimodal-generation/generation \
  -H "Authorization: Bearer $ANYINT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen-image-plus",
    "input": {
      "messages": [
        {
          "role": "user",
          "content": [
            {
              "text": "A premium product hero image for a multi-model AI gateway."
            }
          ]
        }
      ]
    },
    "parameters": {
      "negative_prompt": "",
      "prompt_extend": true,
      "watermark": false,
      "size": "1328*1328"
    }
  }'

Kling v2.1 example

Create a task with kling-v2-1, then poll /transtreams/kling/v1/images/generations/{task_id}.

curl -X POST "https://gateway.api.anyint.ai/transtreams/kling/v1/images/generations" \
  -H "Authorization: Bearer $ANYINT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model_name": "kling-v2-1",
    "model": "kling-v2-1",
    "prompt": "A cyberpunk city street at dusk, premium concept art.",
    "negative_prompt": "blurry, low quality",
    "resolution": "2k",
    "n": 1,
    "aspect_ratio": "3:4"
  }'

Kling V3 Omni image example

The Omni image route is separate from the older Kling image-generation route.

curl -X POST "https://gateway.api.anyint.ai/transtreams/kling/v1/images/omni-image" \
  -H "Authorization: Bearer $ANYINT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model_name": "kling-v3-omni",
    "model": "kling-v3-omni",
    "prompt": "A blue glass cube on a white background, premium product photography.",
    "aspect_ratio": "1:1"
  }'

Implementation advice

Save the original prompt and returned asset URLs together
Treat image generation as a product feature with retries and moderation policies
Validate image size and parameter defaults before exposing them in your UI
Store task IDs for Kling routes so users can resume polling if the page reloads

On this page