CivArchive
    gmic function _kontext Icons become clear - kontext4.2
    NSFW
    Preview 102486219
    Preview 102486222
    Preview 102486230
    Preview 102486240
    Preview 102486262
    Preview 102486278
    Preview 102486291

    Model Basic Information - Basic Information of the Model

    Updated to V4.4, enhanced generalization and support for more styles

    Model Name: gmic function _kontext Icon clarification.

    Model Version: FLUX1.kontext

    Applicable Scenarios: Used for clarifying blurry images of various types of icons and scenes

    Recommended Image Resolution: Determined by the image size

    kontext1.0:Trigger Word: convert to a clear image,

    kontext2.0:convert to gmict hd image

    kontext3.0:clarixgmic++

    kontext4.2:gmic_clearv4

    LORA Weight: 0.8 - 1

    Scheduling Type: SIMple

    Sampling Algorithm: Euler a

    Matching Large Model: flux1-dev-kontext_fp8_scaled.safetensors

    Software: COMFYUI

    Model Author: Game AI Research Institute

    kontext1.0: Text-to-Image Tag Specification: convert to a clear image____style

    kontext2.0: Text-to-Image Tag Specification: convert to gmict hd image, _____style

    kontext3.0: clarixgmic++, _____style(触发词+画风)

    kontext4.2:

    输入图像模糊程度:低 gmic_clearv4,gmic_clearv4 +

    输入图像模糊程度:中 gmic_clearv4,gmic_clearv4 ++

    输入图像模糊程度:较高 gmic_clearv4,gmic_clearv4 +++

    输入图像模糊程度:高(十分模糊) gmic_clearv4,gmic_clearv4 max

    🔗 Model Authorization Address DISCORD: https://discord.gg/njBMYJ7mRF

    QQ Channel (GAMEAI6666): https://pd.qq.com/g/GameAI666

    Description

    模型基本信息 - Basic Information of the Model

    模型名称:gmic function _kontext Icon clarificationV4.2.safetensors

    模型版本: FLUX1.kontext

    适用场景: 用于各类型图标和场景的模糊图变清晰

    推荐图片分辨率:1024PX

    触发词: gmic_clearv4,

    LORA权重0.8 ~ 1

    调度类型: SIMple

    采样算法 :Euler a

    匹配大模型: flux1-dev-kontext_fp8_scaled.safetensors

    软件COMFYUI

    模型作者:游戏AI研究所

    🏷 文生图标签规范:

    输入图像模糊程度:低 gmic_clearv4,gmic_clearv4 +

    输入图像模糊程度:中 gmic_clearv4,gmic_clearv4 ++

    输入图像模糊程度:较高 gmic_clearv4,gmic_clearv4 +++

    输入图像模糊程度:高(十分模糊) gmic_clearv4,gmic_clearv4 max

    🔗 模型授权地址

    DISCORD: https://discord.gg/njBMYJ7mRF

    QQ频道 (GAMEAI6666): https://pd.qq.com/g/GameAI6666

    FAQ

    Comments (3)

    wdadaSep 30, 2025
    CivitAI

    不错的有进步,但是模型有点大

    civitai3463Oct 2, 2025· 4 reactions
    CivitAI

    The term for what this LoRA is doing is "superresolution" and it would be a good idea to mention that term somewhere on this page. Otherwise people searching for superresolution solutions won't find this LoRA. And this LoRA is quite good, so it would be a pity if nobody finds it.

    krigetaNov 8, 2025
    CivitAI

    where is the workflow?

    LORA
    Flux.1 Kontext

    Details

    Downloads
    287
    Platform
    CivitAI
    Platform Status
    Available
    Created
    9/26/2025
    Updated
    5/13/2026
    Deleted
    -
    Trigger Words:
    gmic_clearv4

    Available On (1 platform)

    Same model published on other platforms. May have additional downloads or version variants.