{"uid":"cap_ajtR5f6UcPHexr8j6KCvX","slug":"gpt-image-2-via-xona-agent-medium-quality-76c7ffb6","name":"GPT Image 2 via Xona Agent (Medium Quality)","description":"AI image generation using GPT Image 2 (OpenAI via Replicate). High-quality photorealism with sharp text rendering and strong instruction following. Quality: medium.","url":"https://api.xona-agent.com/image/gpt-image-2","method":"POST","headers":{},"bodySchema":{"type":"object","properties":{"prompt":{"type":"string","description":"Detailed prompt describing the desired image"},"aspect_ratio":{"type":"string","default":"1:1","description":"1:1 (square) — only ratio supported"},"referenceImage":{"type":"array","items":{"type":"string"},"description":"Optional reference image URLs for image-to-image / editing (up to 16)"}}},"responseSchema":null,"example":null,"exampleRequest":null,"tags":["x402"],"displayCostAmount":"0.12","displayCostAsset":"USDC","priceDynamic":false,"priceHint":null,"priceStatus":"priced","priceSource":"registry","requiresHandshake":false,"reviewCount":0,"rating":{"score":"0.00","successRate":"0.00","reviews":0,"stars":null,"state":"unrated"},"availabilityStatus":"unknown","priceObserved":null,"sessionDeposit":null,"pricing":{"kind":"static","summary":"$0.12/call","primary":{"kind":"static","protocol":"x402","network":"base","amountUsd":"0.12","per":"call","confidence":"exact"},"accepted":[{"kind":"static","protocol":"x402","network":"base","amountUsd":"0.12","per":"call","confidence":"exact"}]},"paymentMethods":[{"uid":"pm_OQoXjuqaRm5Lxb9OZw6Xl","protocol":"x402","methodType":"crypto","chain":"base","mode":"charge","costAmount":"0.12","costPer":"request","priority":0,"asset":"EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v","unit":"request","depositMicros":null,"planRef":null}],"brandName":null,"brandSlug":null,"brandBaseUrl":null,"brandDocsUrl":null,"whatItDoes":"Generates high-quality photorealistic images from text prompts using OpenAI's GPT Image 2 model, with optional reference image input for editing or style transfer.","exampleAgentPrompt":"Generate a photorealistic square image of a cozy coffee shop interior at golden hour, warm lighting, with a chalkboard sign that reads 'Morning Brew' above the counter — use GPT Image 2 at medium quality.","exampleUseCases":[{"title":"Marketing visual from brand description","prompt":"Create a photorealistic image of a sleek black sneaker floating against a white gradient background with the text 'LIMITLESS' in bold letters underneath — it's for a product launch campaign."},{"title":"Editorial illustration with text overlay","prompt":"Generate a square image of a futuristic city skyline at dusk with neon signs that clearly read 'Neo Tokyo 2099' — I need the text to be sharp and legible."},{"title":"Reference-based image editing","prompt":"Take this product photo I'm giving you and regenerate it with a tropical beach background instead of the plain white one, keeping the product looking exactly the same."}],"resultDescription":"A generated image file (URL or binary) depicting the scene described in the prompt. The image is square (1:1 aspect ratio), rendered at medium quality using OpenAI's GPT Image 2 model, with strong photorealism, sharp in-image text rendering, and close adherence to the input prompt. If reference images were provided, they influence the output style or content.","failureModes":["Prompt too vague or ambiguous — output may not match intent","Reference image URLs are inaccessible or malformed — endpoint returns an error or ignores them","Content policy violation in prompt — request rejected by OpenAI safety filters","Only 1:1 aspect ratio is supported — other ratios will be ignored or cause errors","Network timeout or Replicate backend latency under heavy load","Payment failure via x402 — request not processed if USDC balance is insufficient"],"whenToPreferThis":"Choose this endpoint when you need high-quality photorealistic images with accurate text rendering and strong prompt-following, especially when using OpenAI's GPT Image 2 model is a priority. It is well-suited for marketing visuals, product mockups, editorial images, or any use case requiring legible in-image text. Prefer this over lower-fidelity models when detail and instruction-following matter. If you need higher quality or more reference images, consider the sibling max-quality endpoint. This endpoint is cost-effective at $0.12 USDC per image.","instructions":null,"reviewSummary":null,"reviewSummaryHighlights":null,"reviewSummaryConcerns":null,"reviewSummaryGeneratedAt":null,"activationCount":0,"lastUsedAt":null,"lastSuccessfullyRanAt":null,"lastHealthCheckAt":"2026-09-15T12:48:28.273Z","isFirstParty":false}