Image Generation API

Generate images using GPT Image (including ChatGPT Images 2.0), Google Nano Banana, CogView-4, or xAI Grok Imagine.

Endpoint

POST https://blockrun.ai/api/v1/images/generations

Request

{
  "model": "google/nano-banana",
  "prompt": "A futuristic city at sunset",
  "size": "1024x1024",
  "n": 1
}

Parameters

Available Models

CogView-4 Sizes

zai/cogview-4 supports flexible sizes (multiples of 16, max 1440×1440):

How it works — hybrid sync/async flow & settlement

This endpoint is hybrid: fast generations complete synchronously, slow ones switch to an async job you poll. The split is purely by elapsed time:

1. POST /v1/images/generations with an x-payment (or PAYMENT-SIGNATURE) header. The gateway verifies the payment authorization (no USDC moves yet) and starts generation.
2. Fast path (≤30s inline window — most models): generation finishes inline. The gateway settles the payment and returns 200 with the standard { created, data: [...] } body below. This is the only moment a fast-path call is charged.
3. Slow path (>30s — e.g. openai/gpt-image-2 under load): the gateway returns 202 with an async job envelope { id, status: "queued", poll_url, price, payment_status: "verified" }. No USDC has moved.
4. GET {poll_url} — poll every 2–5s with an x-payment header signed by the same wallet (a fresh signature works; the gateway enforces wallet binding, not signature byte-equality; with no header it returns a normal x402 402 challenge so clients can re-sign automatically). While the job runs you get 202 (status: queued | in_progress). When it finishes you get 200 with the image URLs, and that is the moment you are charged — settlement happens exactly once, on the first poll that observes status: "completed".

Async job envelope (202)

{
  "id": "img_8f3a…",
  "object": "image.generation.job",
  "status": "queued",
  "model": "openai/gpt-image-2",
  "size": "1024x1024",
  "n": 1,
  "price": { "amount": "0.063000", "currency": "USD" },
  "payment_status": "verified",
  "created": 1706000000,
  "poll_url": "/api/v1/images/generations/img_8f3a…",
  "poll_instructions": "…"
}

status enum: queued → in_progress → completed | failed.

Poll responses

• 202 running: { id, object, status: "queued" | "in_progress", model, payment_status: "verified" }
• 200 completed (charged here): the standard body below plus price: { amount, currency } and payment: { status: "settled", tx_hash, network }, with PAYMENT-RESPONSE and X-Payment-Receipt (on-chain tx hash) headers.
• 200 failed (not charged): { id, object, status: "failed", model, error, payment_status: "not_charged" }

OpenAI-compatible clients: plain OpenAI SDKs don't understand the 202 envelope. Use a model that completes inline, or the official BlockRun SDKs — Go blockrun-llm-go ≥ v0.17.0 handles the hybrid flow transparently and always returns the synchronous {data:[...]} shape.

Response

{
  "created": 1706000000,
  "data": [
    {
      "url": "https://blockrun.ai/api/media/media/images/2026/04/17/<id>.jpg",
      "source_url": "https://imgen.x.ai/xai-imgen/...",
      "backed_up": true,
      "revised_prompt": "..."
    }
  ],
  "price": { "amount": "0.015000", "currency": "USD" },
  "payment": { "status": "settled", "tx_hash": "0x…", "network": "base" }
}

Response Fields

Why both url and source_url? Most providers return temporary URLs (OpenAI image URLs expire in 60 min, xAI Grok Imagine URLs are flagged xai-tmp-*). BlockRun mirrors each generated image to Google Cloud Storage and returns the permanent proxy URL as url.

Examples

Via ClawRouter (recommended for local use)

ClawRouter handles x402 payments automatically. Start it with openclaw gateway start, then call localhost:8402 directly.

curl -X POST http://localhost:8402/v1/images/generations \
  -H "Content-Type: application/json" \
  -d '{"model":"zai/cogview-4","prompt":"a futuristic city at night","size":"1024x1024"}'

Open the returned URL directly:

URL=$(curl -s -X POST http://localhost:8402/v1/images/generations \
  -H "Content-Type: application/json" \
  -d '{"model":"google/nano-banana","prompt":"your prompt here","size":"1024x1024"}' \
  | python3 -c "import sys,json; print(json.load(sys.stdin)['data'][0]['url'])")

open "$URL"   # macOS
xdg-open "$URL"   # Linux

Direct API (cURL)

curl -X POST https://blockrun.ai/api/v1/images/generations \
  -H "Content-Type: application/json" \
  -H "X-Payment: $PAYMENT_HEADER" \
  -d '{
    "model": "zai/cogview-4",
    "prompt": "a minimalist logo for an AI company",
    "size": "1024x1024"
  }'

Python

from blockrun_llm import ImageClient

client = ImageClient()

# CogView-4 (cheapest, supports Chinese prompts)
result = client.generate(
    "一个未来城市的夜景",
    model="zai/cogview-4",
    size="1344x768"
)
print(result.data[0].url)

# Nano Banana (Gemini 2.5 Flash Image)
result = client.generate(
    "A futuristic AI robot",
    model="google/nano-banana",
    size="1024x1024"
)
print(result.data[0].url)

TypeScript

import { ImageClient } from '@blockrun/llm';

const client = new ImageClient();

// CogView-4
const result = await client.generate('a futuristic city at night', {
  model: 'zai/cogview-4',
  size: '1344x768',
});
console.log(result.data[0].url);

// Nano Banana (Gemini 2.5 Flash Image)
const result2 = await client.generate('A futuristic AI robot', {
  model: 'google/nano-banana',
  size: '1024x1024',
});
console.log(result2.data[0].url);

Pricing

Error Codes

Model Selection Guide

OpenAI Compatibility

This endpoint is compatible with OpenAI's Images API:

from openai import OpenAI

client = OpenAI(
    base_url="https://blockrun.ai/api/v1",
    api_key="unused"  # x402 payment handled via header
)

response = client.images.generate(
    model="zai/cogview-4",
    prompt="A futuristic city",
    size="1024x1024"
)
print(response.data[0].url)

Links

• Video Generation
• Image Editing (img2img)
• Music Generation
• nano-banana Skill
• Intelligence Pricing
• Error Handling