Back to model hub

Gemini 3.1 Flash Image Preview

Gemini/gemini-3.1-flash-image-preview
ImageGemini0.5K, 1K, 2K, 4Kgemini-3.1-flash-image-preview

Fast Gemini image generation preview model with multi-resolution output, suitable for creative iteration and UI drafts.

Gemini 3.1 Flash Image Preview — Groups & pricing

Access groups for this site model, same-day runtime metrics, and per-resolution per-call prices from the console.

官KEY-1.5折
BillingPer call
Latency
102.07s
Throughput
10.99tokens/s
Uptime
62.50%
0.5K / call
¥0.05
1K / call
¥0.07
2K / call
¥0.11
4K / call
¥0.16
官KEY-2.5折
BillingPer call
Latency
Throughput
Uptime
0.5K / call
¥0.08
1K / call
¥0.12
2K / call
¥0.18
4K / call
¥0.26
官KEY-3.5折
BillingPer call
Latency
14.45s
Throughput
106.24tokens/s
Uptime
100.00%
0.5K / call
¥0.11
1K / call
¥0.16
2K / call
¥0.25
4K / call
¥0.37
官KEY-4.5折
BillingPer call
Latency
54.18s
Throughput
27.09tokens/s
Uptime
100.00%
0.5K / call
¥0.14
1K / call
¥0.21
2K / call
¥0.32
4K / call
¥0.48
官KEY-5.5折
BillingPer call
Latency
Throughput
Uptime
0.5K / call
¥0.17
1K / call
¥0.26
2K / call
¥0.39
4K / call
¥0.58
官KEY-7.5折
BillingPer call
Latency
Throughput
Uptime
0.5K / call
¥0.24
1K / call
¥0.35
2K / call
¥0.53
4K / call
¥0.79
官KEY-正价
BillingPer call
Latency
133.42s
Throughput
11.54tokens/s
Uptime
100.00%
0.5K / call
¥0.32
1K / call
¥0.47
2K / call
¥0.71
4K / call
¥1.06

Gemini 3.1 Flash Image Preview — Performance

Compare throughput, latency, end-to-end latency, and error rate across channels.

Throughput
Latency
End-to-end latency
Structured output error rate

Gemini 3.1 Flash Image Preview — Uptime

Shows aggregated uptime and least stable provider uptime over time.

Example code & API

Call OpenAI-compatible routes below. Replace the model field with this model’s routing code and paste an API key from the console into the Authorization header.

Traffic is proxied through the unified gateway; each model type maps to a different endpoint (chat, embeddings, images, rerank, speech, etc.), and the snippets switch automatically for the current type.

In production, set sensible timeouts and retries, and keep request identifiers from responses for debugging.

curl -X POST "https://bananarouter.com/v1beta/models/gemini-3.1-flash-image-preview:generateContent" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [
      {
        "parts": [
          { "text": "Generate a futuristic cyberpunk city at night with cinematic lighting and ultra-high detail." }
        ]
      }
    ],
    "generationConfig": {
      "responseModalities": ["IMAGE"],
      "imageConfig": {
        "aspectRatio": "16:9"
      }
    }
  }'