gpt-image-1.5

由 OpenAI 提供图像

使用 GPT Image 1.5 根据提示词生成图像,可走固定价异步渠道或官方生成与编辑接口。

更新于:
输入提示词文本
模型8款可用
返回方式异步结果链接
价格2 个渠道固定价与官方渠道
Young woman sitting on a turquoise van in a flower field beneath balloons and diary doodlesFlower-field van diary示例 01
Twelve-character Breaking Bad-inspired cast poster with distressed typographic masksBreaking Bad cast grid示例 02
Woman seated in a flowing blush couture gown with a dramatic tulle trainBlush couture gown示例 03
Premium mint lemonade advertisement with citrus splashes and floating iceMint lemonade campaign示例 04
生成示例

从提示词到成图

3D Pixar-style expression sheet, cute stylized avatar with wavy dark hair and round glasses, 3x3 grid layout, peeking through torn paper circles, emotions: wink, sunglasses cool, thinking, happy smile, thumbs up, drinking boba tea, friendly wave, laughing, cute pout, different outfits: green sweater, red jacket, yellow hoodie, striped tee, orange shirt, blue top, purple vest, pink cardigan, teal pullover. Soft cinematic lighting, clean white background, vibrant colors, high detail, adorable cartoon character, playful expressions, consistent character design across all poses

A highly realistic low-angle shot of a confident young woman with long, vibrant red hair, holding a watermelon juice can close to the camera in a dramatic perspective. The can appears large in the foreground due to forced perspective, with sharp focus on the product. The woman has fair skin, natural makeup, and a subtle confident expression, looking slightly down toward the camera. She is wearing a fitted blue sleeveless crop top and light gray sweatpants. Her posture is relaxed yet bold, with one arm extended forward holding the can. The background is a smooth, colorful gradient blending blue, orange, and red tones, creating a vibrant summer aesthetic. Bright natural sunlight illuminates the scene with soft shadows and high contrast, giving a clean commercial photography look. The juice can design features a watermelon theme with red and green colors, minimal modern typography, and realistic texture details like condensation droplets. Ultra-detailed, photorealistic, 8K resolution, sharp focus, shallow depth of field, professional product photography, vibrant colors, cinematic lighting. Add text: in small clean white font at the upper right corner.

Avant-garde fashion advertisement, oversized futuristic sunglasses positioned like sculpture, model casually seated on the frame as if furniture, giant word "VISION" behind in bold white typography, powder blue studio background, reflective floor, luxury eyewear campaign aesthetic, ultra-clean layout, editorial magazine styling, 1:1

Classic American brunch plate: avocado toast, poached egg, crispy bacon, fresh orange juice, placed on wooden cafe table, natural window light, messy casual layout, warm beige tone, food blogger photography, slight film grain texture --ar 4:5

A futuristic cyberpunk mega city at night, massive skyscrapers covered with neon lights, flying cars moving between buildings, holographic advertisements floating in the air, rainy streets reflecting colorful lights, advanced AI technology everywhere, cinematic atmosphere, ultra realistic architecture, futuristic Tokyo inspired city, volumetric fog, dramatic lighting, wide angle shot, highly detailed, 8K, masterpiece, unreal engine 5 render

Vibrant lifestyle food advertisement, smiling woman in a bright hot pink blazer sitting inside a colorful trendy café, holding a spoon and eating an acai berry bowl topped with strawberries, blueberries, banana slices, and granola, branded “Berry Loud” jar on wooden table, playful retro typography reading “BERRY LOUD” in large cream bubble letters, tropical café interior with hanging plants, warm natural sunlight, cheerful atmosphere, bold pink and teal color palette, shallow depth of field, cinematic food photography, ultra realistic, high detail, commercial ad campaign style, 4k

Asset type: portrait image for social post Primary request: create a photorealistic travel selfie portrait, using the uploaded portrait photo as the appearance reference for the person. Scene/backdrop: an open alpine meadow under a vivid blue sky, surrounded overhead by hundreds of colorful Tibetan prayer flags arranged in a circular spiral canopy, with distant green hills and bright daylight. Subject: a young woman with the recognizable facial structure, eyes, hairline direction, and natural skin texture from the uploaded portrait photo, wearing a bright cyan outdoor jacket, white hiking pants, a mustard yellow knit beanie, sunglasses resting on the hat, and a backpack. Style/medium: ultra-realistic mobile travel photography with a dynamic action-camera feel, crisp but natural detail, lively social-media adventure portrait. Composition/framing: vertical 3:4 frame, extreme wide-angle perspective, low-to-high selfie composition, one hand reaching toward the camera in the foreground with strong perspective enlargement, face in the midground, prayer flags forming a colorful radial tunnel above her head, energetic candid smile. Lighting/mood: brilliant high-altitude midday sun, visible lens flare near the upper left, sparkling backlight, bright optimistic mood, high clarity, realistic shadows on grass. Color palette: saturated cyan, mustard yellow, red, orange, green, white, and sky blue, clean contrast, sunlit alpine freshness. Textures/retouching: believable fabric texture, natural skin pores, realistic hair strands, clean optical sharpness on the face, shallow foreground hand blur, small falling snowflakes or white petals crossing the lens. Constraints: preserve the person's recognizable appearance from the uploaded portrait photo; no watermark; no logo; no text; no caption; no signature. Avoid: plastic beauty-filter skin, over-smoothed face, distorted hands, extra fingers, warped prayer flags, unreadable markings on clothing, artificial CGI look Negative Prompt: watermark, logo, text, caption, signature, AI label, brand mark, extra fingers, missing fingers, fused fingers, deformed hands, oversized malformed palm, asymmetrical eyes, crossed eyes, warped face, plastic skin, waxy skin, over-smoothed beauty filter, blurry face, low detail, low resolution, heavy compression artifacts, distorted anatomy, duplicate person, bad perspective, bent flag strings, warped flags, fake CGI look, overexposed face, muddy colors

Place the primary chassis or structural platform perfectly centered as the core spine, and position the main body shell separately (left or top), aligned parallel as a reference form while preserving accurate proportions, stance, and geometry of the real vehicle. Design a multi-layered exploded system with components separated across depth (Z-axis) as well as planar spacing, using clear hierarchy: micro hardware (bolts, screws, washers) tightly grouped, functional assemblies moderately spaced, and major systems widely separated. Organize parts into realistic subsystems depending on the vehicle type, including front assembly (suspension, steering, brakes), rear assembly (suspension, drivetrain, differential), chassis/structure (frame, monocoque, subframes), powertrain (engine or motor with slight internal exposure, transmission, intake, exhaust, cooling), electronics (battery, ECU, wiring harness with clean routing), and wheel systems (4 wheels symmetrically placed with tire, rim, and hub separation). Ensure engineering realism by preserving mechanical relationships, aligning shafts and rotational axes, maintaining symmetry across left/right systems, and grouping micro components logically near their assembly points in precise micro-grid arrangements. Avoid random placement. Use a refined, high-end aesthetic: soft neutral gradient background (white to light grey or warm beige), studio-grade product lighting with soft top illumination, subtle directional shadows, and controlled reflections. Apply realistic materials such as brushed metal, carbon fiber, rubber, and glass with slight wear—avoid toy-like finishes. Maintain strong horizontal symmetry with controlled depth asymmetry for visual complexity, and ensure balanced negative space to avoid overcrowding. Render in ultra-realistic quality with razor-sharp edges, macro-level detail, subtle ambient occlusion, and ray-traced shadows. Output in 2K or 4K resolution, 4:5 aspect ratio. No text, no labels, no arrows, no branding overlays, no stylization, no perspective distortion. Style keywords: F1 engineering exploded diagram, luxury automotive technical poster, industrial design breakdown, Apple-style product visualization, ultra-precision mechanical layout, museum-grade presentation.

A cinematic Iceland road trip scene, a lone traveler wearing a rust-colored weatherproof jacket standing beside a rugged black off-road SUV on a vast black sand beach. In the distance, a powerful waterfall cascades down moss-covered volcanic cliffs, surrounded by a dramatic Icelandic landscape. Wind-swept terrain, shallow silver streams reflecting the sky, dark volcanic rocks, low hanging clouds with soft golden sunlight breaking through the horizon. Realistic travel photography, National Geographic documentary style, cinematic color grading, ultra realistic details, natural lighting, shot on a professional full-frame camera, 35mm lens, depth of field, 8K resolution.

模型能力

gpt-image-1.5 能做什么

文生图

通过所选模型的文档接口,用文字生成照片、插画、示意图和宣传素材。

官方渠道输出设置

官方生成接口的画质、尺寸、背景透明度和格式选项因模型而异。

按渠道处理结果

官方生成接口同步返回图像,固定价渠道则提交异步任务,再查询结果。

按模型构造请求

查看所选模型的请求与操作列表,确认支持的字段和返回方式。

计费

对比固定价渠道与官方渠道

根据图像生成或编辑的需要,选择固定价格渠道或官方渠道。

API 请求

按渠道查看请求

切换渠道后,可查看对应的接口地址、请求字段和文档。

非官方渠道

生成

提交固定价 GPT Image 异步任务,再用任务 ID 查询结果。

POSThttps://api.ttapi.io/openai/gpt/generations

请求头

TT-API-KEYstring必填

填写 TTAPI API 密钥。

Content-Typestring必填

使用 application/json。

请求体

promptstring必填

描述想生成的图像。

modelenum选填

可选 gpt-image-2.5、gpt-image-2、gpt-image-2-plus 或 gpt-image-1.5,默认 gpt-image-2。

referImagesstring[]选填

用于引导生成的参考图 URL。

aspect_ratioenum按条件填写

gpt-image-2 支持的宽高比设置。

sizestring按条件填写

gpt-image-2-plus 支持的输出尺寸,格式为 WIDTHxHEIGHT。

hookUrlstring选填

异步任务完成时通知的回调 URL。

查看当前渠道的接入文档
接入示例

查看选中渠道的代码示例

选择渠道和操作,再复制相应的接口地址、请求体或代码示例。

非官方渠道

生成 接入示例

POST https://api.ttapi.io/openai/gpt/generations

curl --request POST \
  --url https://api.ttapi.io/openai/gpt/generations \
  --header "TT-API-KEY: $TTAPI_KEY" \
  --header 'Content-Type: application/json' \
  --data '{"prompt":"minimal product photograph on a warm gray background","model":"gpt-image-1.5","hookUrl":"https://example.com/webhooks/openai-image"}'