模型定價企業
500+ AI 模型 API,全部整合在一個 API 中。就在 CometAPI
模型 API
開發者
快速入門說明文件API 儀表板
公司
關於我們企業
資源
AI模型部落格更新日誌支援
服務條款隱私政策
© 2026 CometAPI · All rights reserved
Home/Models/Midjourney/mj_turbo_upscale_subtle
M

mj_turbo_upscale_subtle

每次請求:$0.168
商業用途
概覽
功能
定價
API

Technical Specifications of mj-turbo-upscale-subtle

SpecificationDetails
Model IDmj-turbo-upscale-subtle
Model familyMidjourney upscale workflow
Primary functionUpscale an already generated Midjourney image while preserving the original composition and look as closely as possible.
Upscale styleSubtle upscale, meaning minimal visual change compared with the source image.
Typical use caseIncreasing image resolution for download, presentation, design review, or higher-resolution asset delivery.
Input dependencyRequires a previously generated Midjourney task/image rather than a brand-new text prompt alone.
Required image selectorimageIndex from 0 to 3 for the chosen image in a four-image Midjourney grid.
API method patternCommon Midjourney upscale APIs use a POST request with a prior taskId and selected image index.
Output characteristicHigher-resolution version of the selected image with limited redesign compared with “creative” upscaling.
Resolution behaviorMidjourney documentation states current subtle/creative upscalers generally double the source image dimensions in supported current versions.
Commercial availabilityCometAPI lists the model as available for commercial use. (cometapi.com)
CometAPI pricingListed at about $0.17 per request on CometAPI at the time of retrieval. (cometapi.com)

What is mj-turbo-upscale-subtle?

mj-turbo-upscale-subtle is CometAPI’s platform identifier for a Midjourney-style upscale operation focused on subtle enhancement rather than heavy reinterpretation. In practice, it is designed for cases where you already have a generated image and want a larger, cleaner version without significantly changing the original framing, textures, or visual intent.

The “subtle” behavior is important: Midjourney’s own documentation distinguishes Subtle from Creative upscaling by explaining that Subtle keeps changes to a minimum and preserves the original image as closely as possible, while Creative may introduce more new details. That makes mj-turbo-upscale-subtle especially suitable for workflows where consistency matters more than reinterpretation.

On CometAPI, this model is exposed as a pay-per-request API offering, which makes it useful for developers building image post-processing pipelines, creative asset tools, automated design systems, or content platforms that need higher-resolution Midjourney outputs through a unified API layer. (cometapi.com)

Main features of mj-turbo-upscale-subtle

  • Subtle visual preservation: The model is aligned with Midjourney’s Subtle upscaler behavior, keeping edits minimal so the upscaled result stays very close to the original image.
  • Higher-resolution output: It is intended to enlarge previously generated images for clearer viewing, sharing, and downstream usage such as print-ready or presentation-friendly assets.
  • Best for approved compositions: Because it avoids major redesign, it works well when a selected generation is already accepted and only needs improved size or finish quality. This is an inference based on the documented “minimal changes” behavior of Subtle upscaling.
  • Task-based workflow: Instead of starting from scratch, the API pattern relies on a previously generated task ID plus a selected image index, making it easy to plug into multi-step generation pipelines.
  • Deterministic image selection: The imageIndex parameter lets developers specify exactly which image from a generated set should be upscaled.
  • Callback-friendly integration: Midjourney upscale API patterns support callback URLs, which helps production systems receive asynchronous completion notifications.
  • Commercial API access through CometAPI: CometAPI lists the model as commercially usable and available with sample code and API resources. (cometapi.com)
  • Simple per-request pricing: The model is presented as a straightforward pay-per-request service, which can be convenient for predictable usage estimation. (cometapi.com)

How to access and integrate mj-turbo-upscale-subtle

Step 1: Sign Up for API Key

To get started, create a CometAPI account and generate your API key from the dashboard. CometAPI provides unified access to many AI models through one API layer, so your key can be used consistently across supported services. After creating the key, store it securely and load it from an environment variable in your application. (cometapi.com)

Step 2: Send Requests to mj-turbo-upscale-subtle API

Use CometAPI's Midjourney-compatible endpoint at POST /mj/submit/action.

curl https://api.cometapi.com/mj/submit/action \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $COMETAPI_API_KEY" \
  -d '{
    "prompt": "a futuristic cityscape at sunset --v 6.1",
    "botType": "MID_JOURNEY",
    "accountFilter": {
      "modes": ["TURBO"]
    }
  }'

Step 3: Retrieve and Verify Results

The API returns a task object with a task ID. Poll GET /mj/task/{task_id}/fetch to check generation status and retrieve the output image URL when the task reaches a terminal state.

mj_turbo_upscale_subtle 的功能

探索 mj_turbo_upscale_subtle 的核心功能,專為提升效能和可用性而設計。了解這些功能如何為您的專案帶來效益並改善使用者體驗。

mj_turbo_upscale_subtle 的定價

探索 mj_turbo_upscale_subtle 的競爭性定價,專為滿足各種預算和使用需求而設計。我們靈活的方案確保您只需為實際使用量付費,讓您能夠隨著需求增長輕鬆擴展。了解 mj_turbo_upscale_subtle 如何在保持成本可控的同時提升您的專案效果。
彗星價格 (USD / M Tokens)官方價格 (USD / M Tokens)折扣
每次請求:$0.168
每次請求:$0.21
-20%

mj_turbo_upscale_subtle 的範例程式碼和 API

存取完整的範例程式碼和 API 資源,以簡化您的 mj_turbo_upscale_subtle 整合流程。我們詳盡的文件提供逐步指引,協助您在專案中充分發揮 mj_turbo_upscale_subtle 的潛力。

更多模型

G

Nano Banana 2

輸入:$0.4/M
輸出:$2.4/M
核心能力概覽:解析度:最高可達 4K(4096×4096),與 Pro 相當。參考圖片一致性:最多支援 14 張參考圖片(10 個物件 + 4 個角色),維持風格與角色一致性。極端寬高比:新增 1:4、4:1、1:8、8:1 比例,適合長圖、海報與橫幅。文字渲染:進階文字生成,適用於資訊圖表與行銷海報版面。搜尋強化:整合 Google Search + Image Search。Grounding:內建思考過程;在生成前會先對複雜提示進行推理。
D

Doubao-Seedance-2-0

每秒:$0.08
Seedance 2.0 是 ByteDance 的新一代多模態影片基礎模型,聚焦於電影級、多鏡頭敘事的影片生成。有別於單鏡頭的文本轉影片示範,Seedance 2.0 著重於基於參考的控制(圖像、短片、音訊)、跨鏡頭的角色與風格一致性,以及原生的聲畫同步——旨在使 AI 影片可用於專業創作與預視化工作流程。
C

Claude Opus 4.7

輸入:$3/M
輸出:$15/M
用於智能體與程式設計的最智能模型
A

Claude Sonnet 4.6

輸入:$2.4/M
輸出:$12/M
Claude Sonnet 4.6 是我們迄今為止最強大的 Sonnet 模型。它對模型在程式設計、電腦操作、長上下文推理、代理規劃、知識工作與設計等方面的能力進行了全面升級。Sonnet 4.6 亦提供 1M 詞元的上下文視窗,目前處於 Beta 階段。
O

GPT-5.4 nano

輸入:$0.16/M
輸出:$1/M
GPT-5.4 nano 專為速度與成本最為關鍵的任務而設計,例如分類、資料擷取、排序與子智能體。
O

GPT-5.4 mini

輸入:$0.6/M
輸出:$3.6/M
GPT-5.4 mini 將 GPT-5.4 的優勢帶入一個更快速、更高效、專為大量工作負載設計的模型。