Gemini generates images (native format)

Use Gemini's native generateContent interface to generate images. Using an image-capable model (such as gemini-2.5-flash-image), the generated image is base64 inlined in parts[].inlineData of the response.

Shares the same endpoint /v1beta/models/{model}:generateContent as the text conversation, except using the image model and declaring IMAGE in generationConfig.responseModalities. When images and text are mixed, TEXT can be returned at the same time.

POST /v1beta/models/{model}:generateContent https://api.covanexus.com

Authorization

Body

Authorization

Authorization BearerAuth

Authorization Bearer <token>

Use Bearer Token authentication: Authorization: Bearer $COVA_NEXUS_API_KEY. Also compatible with Gemini-style x-goog-api-key: $COVA_NEXUS_API_KEY request header.

in: header

Request Body

application/json

contents* array[object]

Generate hints for each of role and parts (text, or inlineData image for graphics)

generationConfig? object

Generate configuration. responseModalities is required for drawing, such as ["TEXT","IMAGE"] or ["IMAGE"]; imageConfig.aspectRatio (such as "1:1", "16:9"), etc. is also supported.

systemInstruction? object

System-level instructions (including parts)

safetySettings? array[object]

Security filtering settings

Response Body

status code	Type
200	application/json
400	application/json
429	application/json

curl -X POST \
  "https://api.covanexus.com/v1beta/models/gemini-2.5-flash-image:generateContent" \
  -H "Authorization: Bearer $COVA_NEXUS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "contents": [
    {"role": "user", "parts": [{"text": "画一只戴贝雷帽的柴犬"}]}
  ],
  "generationConfig": {"responseModalities": ["TEXT", "IMAGE"]}
}'

const res = await fetch(
  "https://api.covanexus.com/v1beta/models/gemini-2.5-flash-image:generateContent",
  {
    method: "POST",
    headers: {
      "Authorization": "Bearer $COVA_NEXUS_API_KEY",
      "Content-Type": "application/json"
    },
    body: JSON.stringify({
      contents: [
        { role: "user", parts: [{ text: "画一只戴贝雷帽的柴犬" }] }
      ],
      generationConfig: { responseModalities: ["TEXT", "IMAGE"] }
    })
  }
);
const data = await res.json();
const img = data.candidates[0].content.parts
  .find(p => p.inlineData)?.inlineData.data; // base64

import base64, requests

url = ("https://api.covanexus.com/v1beta/models/"
       "gemini-2.5-flash-image:generateContent")
resp = requests.post(
    url,
    headers={"Authorization": "Bearer $COVA_NEXUS_API_KEY"},
    json={
        "contents": [
            {"role": "user", "parts": [{"text": "画一只戴贝雷帽的柴犬"}]}
        ],
        "generationConfig": {"responseModalities": ["TEXT", "IMAGE"]}
    }
)
for part in resp.json()["candidates"][0]["content"]["parts"]:
    if "inlineData" in part:
        with open("out.png", "wb") as f:
            f.write(base64.b64decode(part["inlineData"]["data"]))

200

400

429

{
  "candidates": [
    {
      "content": {
        "role": "model",
        "parts": [
          {
            "inlineData": {
              "mimeType": "image.png",
              "data": "<base64>"
            }
          }
        ]
      },
      "finishReason": "STOP",
      "index": 0
    }
  ],
  "usageMetadata": {
    "promptTokenCount": 0,
    "candidatesTokenCount": 0,
    "totalTokenCount": 0
  }
}

BearerAuth