Mindket

QI_2.1_AnyAngle

Newversion 1 · updated

Overview

A LoRA for Qwen Image 2.1 that moves an existing picture to a new camera angle of your choice while keeping its style. It uses a coarse render made from a Gaussian splat or a 3D model.

QI_2.1_AnyAngle is a kit in the Productivity category, listed by Mindket. It is free. This is version 1, last updated on . Download it on this page, a zip of its files. Nothing in it runs by itself.

At a glance

You need
  • Qwen Image 2.1 and the AnyAngle LoRA file
  • Blender, plus a splat or 3D model generator
  • The original image you want to re-angle
Runs in
  • A workflow that can feed two images, an edit prompt and a LoRA into Qwen Image 2.1
  • Blender, to place a new camera and render the coarse view
  • A Gaussian splat or 3D model generator, such as Tripo Splat or Trellis2
  • Windows, macOS or Linux with Python, to download the files
Start with
pip install -U huggingface_hub
hf download lilylilith/QI_2.1_AnyAngle

Platform

WindowsOperating system
macOSOperating system
LinuxOperating system

About

What it is

AnyAngle is a LoRA for Qwen Image 2.1 that changes the camera angle of a picture you already have. Image generators often miss the angle you ask for, and editing models tend to offer only fixed turns, with the style drifting on big moves. AnyAngle takes a coarse render of the scene from the angle you want and transfers that angle onto your original image, keeping the style consistent.

Who it suits

It suits artists and image makers who want an exact new viewpoint on a picture and are comfortable adding Blender to their workflow.

IndustryWho uses itHow they use itExample
IllustrationDigital artistsTurn a finished picture to a new viewpoint and keep its lookShow a character drawing from a higher angle
Film and animationStoryboard artistsPlan shots from several camera anglesTry a turbo version for fast shot planning
Game and 3D artBlender usersPair a Blender render with the original imageRender a coarse view, then restyle it to match the source

If your picture is low resolution or its faces are unclear, results may suffer, so you may prefer another tool.

Where to use it

  • A workflow that can feed two images, an edit prompt and a LoRA into Qwen Image 2.1
  • Blender, to place a new camera and render the coarse view
  • A Gaussian splat or 3D model generator, such as Tripo Splat or Trellis2
  • Windows, macOS or Linux with Python, to download the files
Use caseWhoWhat they get back
New camera angleArtistsYour original image redrawn from the angle you set in Blender
Shot planningStoryboard artistsQuick views of one scene from several angles, using a turbo model
Style-safe turningIllustratorsA turned view that keeps the look of the source image

How to set it up

What you need first:

  • Qwen Image 2.1 and the AnyAngle LoRA file
  • Blender, plus a splat or 3D model generator
  • The original image you want to re-angle
  1. Make a Gaussian splat or a 3D model from your original image.
  2. Import it into Blender, add a camera at the angle you want, and render the coarse image.
  3. Download the LoRA and load it in Qwen Image 2.1 with its strength set to 1.
pip install -U huggingface_hub
hf download lilylilith/QI_2.1_AnyAngle

How to use it

Give Qwen Image 2.1 two images, the original and the coarse render, connected the way the card's picture of the workflow shows. Then use the edit prompt below.

Change the camera angle from <image2> to <image1>.
Settings the card recommends
SettingValue
LoRA strength1
CFG3.0
Steps20 or more
ToolWhat it does
Splat or 3D generatorTurns your image into a scene you can view from other angles
BlenderHolds the new camera and renders the coarse view
Qwen Image 2.1 with AnyAngleMoves the new angle onto your original image

Example

You have a front view of a character. You turn it into a splat, place a Blender camera above and to the side, and render. Qwen Image 2.1 with AnyAngle then redraws the original from that angle in the same style.

For newcomers and experts

A good first try is one picture with a simple subject and a modest camera move. Check the coarse render first, because the final result follows it.

For experts, the LoRA was trained on pairs. Each pair holds an anchor picture and a frame of the same scene from another angle, called the target. The target becomes a splat or 3D model, and its coarse render serves as the control image. The data mixes Blender renders in many styles with turn-around videos made by MiniMax H3 for digital illustrations, where the subject stays still. Training lasted a few thousand steps. Because the real Blender renders contain nothing invented, the style stays aligned.

  1. Your image becomes a splat or a 3D model
  2. Import it into Blender and set a new camera angle
  3. Render the coarse image from that angle
  4. Qwen Image 2.1 with AnyAngle renders the final picture

Good to know

  • Limits: weak splats or models, extreme anatomy, unclear faces and low resolution can cause odd layouts or malformed faces.
  • Speed: a turbo LoRA or a turbo Qwen Image 2.1 model lowers latency, with a slight drop in quality.
  • Base model: Qwen Image 2.1.
  • Tools: Tripo Splat or Trellis2 are the ones the card used, and it says others would work.
  • License: apache-2.0
  • Cost to run: no API key or paid service named. The splat and 3D tools you pick may have their own terms.
  • Data it reaches: it downloads files from Hugging Face. The splat and 3D tools you pick may send your image elsewhere, so check them.
  • Not for: pictures with badly formed faces or anatomy, or tiny images.

README and files

README

QI_2.1_AnyAngle

A LoRA for Qwen Image 2.1 that moves an existing picture to a new camera angle of your choice while keeping its style. It uses a coarse render made from a Gaussian splat or a 3D model.

Made by lilylilith on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face.

About

What it is: AnyAngle is a LoRA for Qwen Image 2.1 that changes the camera angle of a picture you already have. Image generators often miss the angle you ask for, and editing models tend to offer only fixed turns, with the style drifting on big moves. AnyAngle takes a coarse render of the scene from the angle you want and transfers that angle onto your original image, keeping the style consistent.

Who it suits: It suits artists and image makers who want an exact new viewpoint on a picture and are comfortable adding Blender to their workflow.

Industry | Who uses it | How they use it | Example
Illustration | Digital artists | Turn a finished picture to a new viewpoint and keep its look | Show a character drawing from a higher angle
Film and animation | Storyboard artists | Plan shots from several camera angles | Try a turbo version for fast shot planning
Game and 3D art | Blender users | Pair a Blender render with the original image | Render a coarse view, then restyle it to match the source

If your picture is low resolution or its faces are unclear, results may suffer, so you may prefer another tool.

Where to use it:

  • A workflow that can feed two images, an edit prompt and a LoRA into Qwen Image 2.1
  • Blender, to place a new camera and render the coarse view
  • A Gaussian splat or 3D model generator, such as Tripo Splat or Trellis2
  • Windows, macOS or Linux with Python, to download the files
Use case | Who | What they get back
New camera angle | Artists | Your original image redrawn from the angle you set in Blender
Shot planning | Storyboard artists | Quick views of one scene from several angles, using a turbo model
Style-safe turning | Illustrators | A turned view that keeps the look of the source image

How to set it up: What you need first:

  • Qwen Image 2.1 and the AnyAngle LoRA file
  • Blender, plus a splat or 3D model generator
  • The original image you want to re-angle
  1. Make a Gaussian splat or a 3D model from your original image.
  2. Import it into Blender, add a camera at the angle you want, and render the coarse image.
  3. Download the LoRA and load it in Qwen Image 2.1 with its strength set to 1.
pip install -U huggingface_hub
hf download lilylilith/QI_2.1_AnyAngle

How to use it: Give Qwen Image 2.1 two images, the original and the coarse render, connected the way the card's picture of the workflow shows. Then use the edit prompt below.

Change the camera angle from <image2> to <image1>.
Title: Settings the card recommends
Setting | Value
LoRA strength | 1
CFG | 3.0
Steps | 20 or more
Tool | What it does
Splat or 3D generator | Turns your image into a scene you can view from other angles
Blender | Holds the new camera and renders the coarse view
Qwen Image 2.1 with AnyAngle | Moves the new angle onto your original image

Example: You have a front view of a character. You turn it into a splat, place a Blender camera above and to the side, and render. Qwen Image 2.1 with AnyAngle then redraws the original from that angle in the same style.

For newcomers and experts: A good first try is one picture with a simple subject and a modest camera move. Check the coarse render first, because the final result follows it.

For experts, the LoRA was trained on pairs. Each pair holds an anchor picture and a frame of the same scene from another angle, called the target. The target becomes a splat or 3D model, and its coarse render serves as the control image. The data mixes Blender renders in many styles with turn-around videos made by MiniMax H3 for digital illustrations, where the subject stays still. Training lasted a few thousand steps. Because the real Blender renders contain nothing invented, the style stays aligned.

Your image becomes a splat or a 3D model
Import it into Blender and set a new camera angle
Render the coarse image from that angle
Qwen Image 2.1 with AnyAngle renders the final picture

Good to know:

  • Limits: weak splats or models, extreme anatomy, unclear faces and low resolution can cause odd layouts or malformed faces.
  • Speed: a turbo LoRA or a turbo Qwen Image 2.1 model lowers latency, with a slight drop in quality.
  • Base model: Qwen Image 2.1.
  • Tools: Tripo Splat or Trellis2 are the ones the card used, and it says others would work.
  • License: apache-2.0
  • Cost to run: no API key or paid service named. The splat and 3D tools you pick may have their own terms.
  • Data it reaches: it downloads files from Hugging Face. The splat and 3D tools you pick may send your image elsewhere, so check them.
  • Not for: pictures with badly formed faces or anatomy, or tiny images.

On Hugging Face, 13 November 2026

  • Task: image to image
  • Library: diffusers
  • Downloads in the last 30 days: 13,198
  • Likes: 189
  • License: apache-2.0
  • Based on: Qwen/Qwen-Image-2.1
  • Tags: diffusers, lora, qwen-image-2.1, image-to-image

Links

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download lilylilith/QI_2.1_AnyAngle

Then follow the usage in its model card. Nothing in this kit runs by itself.

Files
README.md  6 KB

1 file, 6 KB, downloaded as a zip.

What’s new

  • Version 1 · 2026-11-13

Compare with similar

Listed Hugging Face models in the same category that share the most tags with this one.

ProductDownloads, last 30 daysLikesLicenseLast change
QI_2.1_AnyAngle This page13,198189apache-2.0
Qwen-Image-2.1-Next-Scene-LoRAShares qwen-image-2.1 · image-to-image · lora4,58171apache-2.0
Qwen-Image-2.1-Multiple-Angles-LoRAShares qwen-image-2.1 · image-to-image · lora3,505110apache-2.0
MOSS-Image depth control LoRAShares image-to-image · lora · diffusers2,0116apache-2.0
Moss-Image-Edit LoRA for ComfyUIShares image-to-image · lora018gpl-3.0
Qwen-Image-2.1-LoRAsShares lora · diffusers28,839138apache-2.0
Numbers as Mindket last read them, from 2026-11-12 to 2026-11-13.

Reviews

No ratings yet

Write a review

No reviews yet.

More from Mindket

See all from Mindket

Huashu Design

alchaincyf · Skill

Turns a one-line request in your AI agent into a finished design: a launch animation, a clickable app prototype, an editable slide deck, an infographic or a se…

New29.4k

Guizang PPT Skill

op7418 · Skill

Skill that makes single-file HTML slide decks, slide illustrations and platform covers in two visual styles, with a presenter mode for rehearsal. For speakers…

New29.8k

Claude Code Skills & Plugins

alirezarezvani · Skill

Adds 411 ready-made skills for Claude Code and other coding agents, grouped by role from engineering and product to marketing, finance and compliance. Installs…

New29.8k

Similar in Productivity

See all in Productivity

Frontend Slides

zarazhangrui · Skill

Builds HTML presentations from scratch or from a PowerPoint file and shows visual style previews to pick from, for people who want web slides with no CSS or Ja…

New32.3k

IPTV

iptv-org · Kit

Collects publicly available live TV channels from many countries into playlists that any video player with live streaming can open. Made for viewers and for de…

New144k

upscaler

uwg · Kit

Holds a set of image upscaling models gathered from the OpenModelDB community, for people who want to enlarge low-resolution pictures with AI on their own comp…

New