Skip to content

Image Model Matrix ​

This page summarizes image models currently listed on the pricing page, their recommended endpoints, input formats and result handling. Check the console, pricing page and GET /v1/models for actual model availability, sizes, counts and prices.

Model summary tables ​

Model family.Representative ModelRecommended entranceUsual ModeQuery Entry
GPT Imagegpt-image-2, gpt-image-2-pro, gpt-image-2-vip/v1/images/generations, /v1/images/edits, /v1/images/generations/async, /v1/images/edits/asyncText-to-image and image editingSynchronous results return directly; query asynchronous tasks at /v1/images/tasks/{task_id}
Seedream / Wan asynchronous pictureseedream-5.0, seedream-4.5, wan2.7-image-pro/v1/images/generations/asynctext-to-image, image-to-image/v1/images/tasks/{task_id}
Gemini Nanogemini-3.1-flash-image-preview, gemini-3-pro-image-preview, nano-banana-2, nano-banana-proSynchronous /v1beta/models/{model}:generateContent; /v1/images/generations/async when asynchronous access is enabledGemini-format image generation and unified asynchronous tasksSynchronous results return directly; query asynchronous tasks at /v1/images/tasks/{task_id}
Doubao Seedreamdoubao-seedream-4-5-251128, doubao-seedream-5-0-260128/v1/images/generationstext-to-image, Creativesynchronous returns, no task queries
Grok Imaginegrok-imagine-image, grok-imagine-image-pro/v1/images/generationstext-to-image, Creativesynchronous returns, no task queries
Ideogram imagesideogram-*/ideogram/generate, /ideogram/remix, /ideogram/upscale, /ideogram/describeText-to-image, remix, upscale and image descriptionsFollow the actual response
Midjourney taskmj-*/mj/submit/*Imagine, mixed, described, action task/mj/task/{id}/fetch or /mj/task/list-by-condition

Photo Open Model ​

ModelsCapacityInputOutput
gpt-image-2Picture GenerationPromptPhoto URL or image content, as per API Reference
gpt-image-2-proHigh-quality photo generation and editingPrompt, Photo FilesPhoto URL or image content, as per API Reference
gpt-image-2-vipGPT Image Fast or High Priority ModePromptPhoto URL or image content, as per API Reference
doubao-seedream-4-5-251128Seedream Image GenerationPromptPhoto URL or image content, as per API Reference
doubao-seedream-5-0-260128Seedream Image GenerationPromptPhoto URL or image content, as per API Reference
seedream-5.0asynchronous Photo Generation and image-to-imagePrompt, Photo URLImage task result URL
seedream-4.5asynchronous Photo Generation and image-to-imagePrompt, Photo URLImage task result URL
wan2.7-image-proasynchronous Photo Generation and image-to-imagePrompt, Photo URLImage task result URL
gemini-3.1-flash-image-previewGemini Image GenerationGemini contents[]Gemini-compatible response
gemini-3-pro-image-previewGemini image generation; unified asynchronous tasks when enabledGemini contents[], or prompt on /v1/images/generations/asyncGemini-compatible response or image-task result URL
nano-banana-2Gemini Nano Image GenerationGemini contents[]Gemini-compatible response
nano-banana-proGemini Nano Image GenerationGemini contents[]Gemini-compatible response
grok-imagine-imageGrok Imagine Image GenerationPromptPhoto URL or image content, as per API Reference
grok-imagine-image-proGrok Imagine Pro Image GenerationPromptPhoto URL or image content, as per API Reference
ideogram-*Text-to-image, remix, upscale and image descriptionsPrompt and image URLImage result or descriptive text, according to the response
mj-*Imagine, mixed, described, action taskPrompt, Photo URL, action parameterstask status, progress, results picture and button action

Enter media assets Mode ​

Input methodSuitable usesCall the asset API first?
Prompttext-to-image, poster, styled creative map.No need. Pass it to prompt.
Multipart Photo FilePicture editing and photo variants.Not required. Directly to /v1/images/edits or /v1/images/variations.
Clients' own public network URLIdeogram, Midjourney or model supports URL input.Not required. Ensure that URLs are stable and accessible.
MoonNexAI OSS Test URLFor local document association only, there is no guarantee of long-term preservation or continued availability.Call /v1/assets/uploads and use mode=oss; for official business, switch to the customer-owned public network URL.

Generate Mode Support ​

Model family.text-to-imagePicture EditorPicture variantMixed ChartZoom in.Picture DescriptionAsynchronous tasks
GPT ImageSupportAvailable when supportedBased on model descriptionNot applicableNot applicableNot applicableSupport
Seedream / Wan asynchronous pictureSupportimage-to-image AvailableNot applicableNot applicableNot applicableNot applicableSupport
Gemini NanoSupportBased on model descriptionNot applicableNot applicableNot applicableBased on Gemini CapabilitySupport when asynchronous capability is in place
Doubao SeedreamSupportBased on model descriptionNot applicableNot applicableNot applicableNot applicableNot applicable
Grok ImagineSupportBased on model descriptionNot applicableNot applicableNot applicableNot applicableNot applicable
Ideogram imagesSupportedAvailable for remixNot applicableSupportedSupportedSupportedFollow the actual response
Midjourney taskSupportAction task availableAction task availableSupportSupportSupportSupport

Common field ​

fieldPurpose
modelPhoto model name, required for General Image API.
promptPhoto description, with recommendations for subject, scene, style, aspect ratio and use.
imageMultipart photo file for photo editing and transformation.
image_urlPhoto URL, used for Ideogram, picture description or compatibility URL input.
sizeOutput sizes, e.g. 1024x1024, the exact value is based on the range of model support.
aspect_ratioPicture aspect ratio, e.g. 1:1, 16:9, 9:16.
nGenerates the quantity.
response_formatreturns the format, which is based on API Reference.
task_idPicture asynchronous tasks ID, for GET /v1/images/tasks/{task_id}.
notifyHookThe address of callback upon completion of Midjourney task.
customIdMidjourney action task uses buttons or action IDs.

Entry Selection ​

RequirementsRecommended Model FamilyRecommended PagesCommon field
Generate pictures from textGPT Image, Doubao Seedream, Grok ImaginePicture Generationmodel, prompt, size, n
Time-consuming pictures generation or editingGPT Image, Seedream / Wan asynchronous pictures, Gemini image model with asynchronous capabilitiesPicture Step Taskmodel, prompt, size, resolution, task_id
Use seedream-5.0, seedream-4.5, wan2.7-image-proSeedream / Wan asynchronous pictureSeedream / Wan Step Picture Modelmodel, input.messages, parameters.size
Upload original maps and modify them by descriptionGPT ImagePicture Editorimage, model, prompt, size
Generate pictures in Gemini formatGemini NanoGemini Nano Synchronize Callcontents[], parts[].text
Posters, packaging or text layoutIdeogram ImageIdeogram text-to-imageprompt, aspect_ratio
Based on the image to continue mixing or zooming inIdeogram ImageIdeogram Imageimage_url, prompt, aspect_ratio
Submit high-stylish asynchronous pictures taskMidjourney taskMidjourney Imagineprompt, notifyHook, customId

Agreement description ​

MoonNexAI /v1/images/* uses OpenAI Images-compatible endpoints for models such as gpt-image-*, doubao-seedream-* and grok-imagine-*. Synchronous calls wait for completion. Asynchronous calls return a task ID queried at /v1/images/tasks/{task_id}. seedream-5.0, seedream-4.5, wan2.7-image-pro and enabled gemini-3-pro-image-preview accounts can use the asynchronous endpoint. Synchronous Gemini-compatible requests use /v1beta/models/{model}:generateContent. Check the console, pricing page and GET /v1/models for editing, variations, sizes, resolutions and output formats. Asynchronous access is not guaranteed for every API key group.

/v1/chat/completions is used for chat completion and is not a recommended photo generation portal for gpt-image-*. If you need a picture, call the Images API by the recommended entry in the model matrix.

Relevant Pages ​