Skip to page content
SAVRN
Search Plan a project

Open-weight model · Image to image

Qwen-Image-2.1-Next-Scene-LoRA

by AK akhaliq/Qwen-Image-2.1-Next-Scene-LoRA

Qwen-Image-2.1-Next-Scene-LoRA is an open-weight model for image to image from AK, released under Apache License 2.0. Its published files total 3.5 GB.

Feed a frame, direct the next shot. This LoRA teaches Qwen-Image-2.1 cinematic scene continuation: it takes an input image and a director-style instruction and generates the next scene — camera moves, reveals, lighting shifts — while keeping the world…

Parameters—
Context—
Weights3.5 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads—

Model Card

By AK, published under apache-2.0, revision d05eeb04ce08.

Feed a frame, direct the next shot. This LoRA teaches Qwen-Image-2.1 cinematic scene continuation: it takes an input image and a director-style instruction and generates the next scene — camera moves, reveals, lighting shifts — while keeping the world, characters and style consistent. Chain outputs into storyboards. This is the Qwen-Image-2.1 port of the next-scene format pioneered by (313K downloads on the 2509 edit model) — retrained natively on the newest Qwen image base, where this capability did not exist yet. Companion piece to (camera control) — together they cover where the camera is and where the story goes next. ComfyUI: load in any Qwen-Image-2.1 workflow with a LoRA node…

Read AK's full model card

Feed a frame, direct the next shot. This LoRA teaches Qwen-Image-2.1 cinematic scene continuation: it takes an input image and a director-style instruction and generates the next scene — camera moves, reveals, lighting shifts — while keeping the world, characters and style consistent. Chain outputs into storyboards.

This is the Qwen-Image-2.1 port of the next-scene format pioneered by lovis93/next-scene-qwen-image-lora-2509 (313K downloads on the 2509 edit model) — retrained natively on the newest Qwen image base, where this capability did not exist yet. Companion piece to akhaliq/Qwen-Image-2.1-Multiple-Angles-LoRA (camera control) — together they cover where the camera is and where the story goes next.

Usage

from diffusers import QwenImage21Pipeline
import torch

pipe = QwenImage21Pipeline.from_pretrained("Qwen/Qwen-Image-2.1", torch_dtype=torch.bfloat16)
pipe.load_lora_weights(
    "akhaliq/Qwen-Image-2.1-Next-Scene-LoRA",
    weight_name="checkpoints/qwen21_next_scene_v1_000001500.safetensors",
)

image = pipe(
    prompt="Next Scene: The camera pulls back to a sweeping aerial view, "
           "revealing the fleet behind the cliffs. Soft morning light, atmospheric depth.",
    image=input_image,          # your current frame
    num_inference_steps=28,
    true_cfg_scale=1.0,
).images[0]

ComfyUI: load in any Qwen-Image-2.1 workflow with a LoRA node (strength 0.7–0.9), pass your image as the edit/reference input.

Prompt format: start with Next Scene: and lead with the camera direction:

Next Scene: The camera tracks forward and tilts down as rain begins to streak the lens.
Next Scene: Cut to a low-angle shot; sunlight breaks through the clouds behind her.
Next Scene: The camera pans right, revealing the fleet massing behind the ridge.

Pro tips - Begin with the camera move, then the subject/lighting change — the model was trained on captions written this way. - Chain generations (output → input) to build sequential storyboards. - If the output drifts toward copying the input, lower the LoRA strength toward 0.7.

Which checkpoint

Checkpoint Verdict
checkpoints/qwen21_next_scene_v1_000001500.safetensors Recommended. Balanced instruction-following vs identity preservation (mid-training peak, same regime as prior Qwen-Image-2.1 LoRA runs).
…_000001000.safetensors Lighter touch — try if 1500 over-stylizes your subject.
…_000002000.safetensors / …_000002500.safetensors Stronger-baked scene transitions; earlier checkpoints hold identity better.
qwen21_next_scene_v1.safetensors = step 2500 (final).

Per-250-step sample grids from the training run are in checkpoints/samples/ so the training curve is inspectable end to end.

Training

  • Base: Qwen/Qwen-Image-2.1 (unified generate + edit), arch: qwen_image_2
  • Trainer: ostris/ai-toolkit — rank 64 / alpha 64, adamw8bit, lr 1e-4, weighted timesteps, caption dropout 0.05, 512², 2,500 steps, uint3-quantized base, bf16 training
  • Data: 1,144 scene-progression pairs — consecutive frames from the Blender open films Sintel and Tears of Steel (© Blender Foundation, CC-BY, via KevinConnorLee/Sintel (MPI-Sintel frames) and the Blender mirrors). Captions are director-style "Next Scene: …" instructions written by Qwen3-VL-4B over each (frame → next frame) pair. Published: pairs · captioned training set
  • Dashboard: trackio training curves
  • Demo: akhaliq/qwen21-next-scene

Scope & limits

  • Trained on cinematic film frames (one animated, one live-action film) — it generalizes to photographic inputs, but the strongest results are on filmic scenes, landscapes and establishing shots; static portraits are out of scope by design.
  • Trained at 512²; larger canvases inherit the behavior but were not trained on.
  • This is a v1: 1.1K pairs from two films. A v2 with broader footage is the planned upgrade.
  • Checkpoint selection here follows the mid-training-peak heuristic from prior Qwen-Image-2.1 LoRA runs on this trainer; judge checkpoints/samples/ grids and pick your own if you prefer a stronger or lighter bake.

Identity and Version

Repository
akhaliq/Qwen-Image-2.1-Next-Scene-LoRA
Publisher
AK
Task
Image to image
Modality
Image
Library
diffusers
Parameters
Not stated by the source
Languages
ai-video
Revision
d05eeb04ce08f4935884175845424f5663b833bf
First published
2026-10-07
Last updated
2026-10-08

Files and Weights

66 files, 3.5 GB in total. The weights are 11 files totalling 3.5 GB in pt, safetensors.

Weights11 files · 3.5 GB
Configuration1 file · 1.6 KB
Documentation1 file · 5.1 KB
Other52 files · 1.3 MB
Repository1 file · 1.5 KB
Every file
FileTypeSizeSHA-256
checkpoints/optimizer.ptWeights324.2 MB 01ebbe5f435a
checkpoints/qwen21_next_scene_v1.safetensorsWeights318.8 MB 8a87032dd1d2
checkpoints/qwen21_next_scene_v1_000000250.safetensorsWeights318.8 MB 85fa6fa7913d
checkpoints/qwen21_next_scene_v1_000000500.safetensorsWeights318.8 MB 58af76786d7c
checkpoints/qwen21_next_scene_v1_000000750.safetensorsWeights318.8 MB 38db0ecbd29c
checkpoints/qwen21_next_scene_v1_000001000.safetensorsWeights318.8 MB 486d9c493f4b
checkpoints/qwen21_next_scene_v1_000001250.safetensorsWeights318.8 MB 248fcc385544
checkpoints/qwen21_next_scene_v1_000001500.safetensorsWeights318.8 MB c7cbaba18317
checkpoints/qwen21_next_scene_v1_000001750.safetensorsWeights318.8 MB df42d17d324b
checkpoints/qwen21_next_scene_v1_000002000.safetensorsWeights318.8 MB f76fa34a7c85
checkpoints/qwen21_next_scene_v1_000002250.safetensorsWeights318.8 MB 4a8183bbeea6
checkpoints/config.yamlConfiguration1.6 KB —
README.mdDocumentation5.1 KB —
checkpoints/samples/.thumbs/1791419300504__000000000_0.jpg.jpgOther25.1 KB —
checkpoints/samples/.thumbs/1791419364011__000000000_1.jpg.jpgOther26.5 KB —
checkpoints/samples/.thumbs/1791419804659__000000100_0.jpg.jpgOther23.0 KB —
checkpoints/samples/.thumbs/1791419868128__000000100_1.jpg.jpgOther14.8 KB —
checkpoints/samples/.thumbs/1791420916230__000000000_0.jpg.jpgOther27.2 KB —
checkpoints/samples/.thumbs/1791420967271__000000000_1.jpg.jpgOther24.3 KB —
checkpoints/samples/.thumbs/1791421848998__000000250_0.jpg.jpgOther20.8 KB —
checkpoints/samples/.thumbs/1791421900044__000000250_1.jpg.jpgOther17.1 KB —
checkpoints/samples/.thumbs/1791422767013__000000500_0.jpg.jpgOther23.5 KB —
checkpoints/samples/.thumbs/1791422818075__000000500_1.jpg.jpgOther16.7 KB —
checkpoints/samples/.thumbs/1791423695176__000000750_0.jpg.jpgOther21.7 KB —
checkpoints/samples/.thumbs/1791423746239__000000750_1.jpg.jpgOther20.1 KB —
checkpoints/samples/.thumbs/1791424614825__000001000_0.jpg.jpgOther22.9 KB —
checkpoints/samples/.thumbs/1791424665914__000001000_1.jpg.jpgOther25.2 KB —
checkpoints/samples/.thumbs/1791425518578__000001250_0.jpg.jpgOther23.4 KB —
checkpoints/samples/.thumbs/1791425569648__000001250_1.jpg.jpgOther22.8 KB —
checkpoints/samples/.thumbs/1791426438214__000001500_0.jpg.jpgOther21.5 KB —
checkpoints/samples/.thumbs/1791426489283__000001500_1.jpg.jpgOther22.2 KB —
checkpoints/samples/.thumbs/1791427361171__000001750_0.jpg.jpgOther19.4 KB —
checkpoints/samples/.thumbs/1791427412234__000001750_1.jpg.jpgOther24.3 KB —
checkpoints/samples/.thumbs/1791428283616__000002000_0.jpg.jpgOther11.4 KB —
checkpoints/samples/.thumbs/1791428334670__000002000_1.jpg.jpgOther17.8 KB —
checkpoints/samples/.thumbs/1791429208351__000002250_0.jpg.jpgOther25.0 KB —
checkpoints/samples/.thumbs/1791429259435__000002250_1.jpg.jpgOther20.7 KB —
checkpoints/samples/.thumbs/1791430125430__000002500_0.jpg.jpgOther25.2 KB —
checkpoints/samples/.thumbs/1791430176507__000002500_1.jpg.jpgOther21.6 KB —
checkpoints/samples/1791419300504__000000000_0.jpgOther35.7 KB —
checkpoints/samples/1791419364011__000000000_1.jpgOther38.3 KB —
checkpoints/samples/1791419804659__000000100_0.jpgOther30.9 KB —
checkpoints/samples/1791419868128__000000100_1.jpgOther19.1 KB —
checkpoints/samples/1791420916230__000000000_0.jpgOther38.3 KB —
checkpoints/samples/1791420967271__000000000_1.jpgOther36.1 KB —
checkpoints/samples/1791421848998__000000250_0.jpgOther26.3 KB —
checkpoints/samples/1791421900044__000000250_1.jpgOther21.8 KB —
checkpoints/samples/1791422767013__000000500_0.jpgOther29.6 KB —
checkpoints/samples/1791422818075__000000500_1.jpgOther20.9 KB —
checkpoints/samples/1791423695176__000000750_0.jpgOther27.5 KB —
checkpoints/samples/1791423746239__000000750_1.jpgOther26.0 KB —
checkpoints/samples/1791424614825__000001000_0.jpgOther29.7 KB —
checkpoints/samples/1791424665914__000001000_1.jpgOther34.9 KB —
checkpoints/samples/1791425518578__000001250_0.jpgOther29.4 KB —
checkpoints/samples/1791425569648__000001250_1.jpgOther29.6 KB —
checkpoints/samples/1791426438214__000001500_0.jpgOther27.5 KB —
checkpoints/samples/1791426489283__000001500_1.jpgOther29.0 KB —
checkpoints/samples/1791427361171__000001750_0.jpgOther25.7 KB —
checkpoints/samples/1791427412234__000001750_1.jpgOther33.0 KB —
checkpoints/samples/1791428283616__000002000_0.jpgOther16.6 KB —
checkpoints/samples/1791428334670__000002000_1.jpgOther24.4 KB —
checkpoints/samples/1791429208351__000002250_0.jpgOther31.9 KB —
checkpoints/samples/1791429259435__000002250_1.jpgOther27.9 KB —
checkpoints/samples/1791430125430__000002500_0.jpgOther32.6 KB —
checkpoints/samples/1791430176507__000002500_1.jpgOther27.7 KB —
.gitattributesRepository1.5 KB —

License and Download

License
apache-2.0
Access
Open weights, no gate
Download size
3.5 GB
Download from AK

Released by AK through its official repository on Hugging Face. Read the license.

Built From

  • Adapter of Qwen/Qwen-Image-2.1
  • Derived from Qwen/Qwen-Image-2.1

Memory Requirements

PrecisionWeights in memory
As published3.5 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Questions About Qwen-Image-2.1-Next-Scene-LoRA

Can I use Qwen-Image-2.1-Next-Scene-LoRA commercially?

Yes. Qwen-Image-2.1-Next-Scene-LoRA is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.

Similar Models

Model · Image to image

FLUX.2-klein-4B-GGUF

Unsloth AI

This is a GGUF quantized version of FLUX.2-klein-4B. unsloth/FLUX.2-klein-4B-GGUF uses Unsloth Dynamic 2.0 methodology for SOTA performance. - Important layers are upcasted to higher precision. - Uses tooling from ComfyUI-GGUF by city96. The FLUX.2 [klein] model family are our fastest image models to date. FLUX.2 [klein] unifies generation and editing in a single compact architecture, delivering state-of-the-art quality with end-to-end inference in as low as under a second. Built for applications that require real-time image generation without sacrificing quality, and runs on consumer hardware, with as little as 13GB VRAM. FLUX.2 [klein] 4B is a 4 billion parameter rectified flow…

Open weights apache-2.0 ggml
View model

Model · Image to image

FLUX.2-klein-4b-fp8

Black Forest Labs

FLUX.2 [klein] 4B is a 4 billion parameter rectified flow transformer capable of generating images from text descriptions and supports multi-reference editing capabilities. For more information, please read our blog post. This repository holds an FP8 version of FLUX.2 [klein] 4B. The main repository of this model (full BF16 weights) can be found here. Limitations - This model is not intended or able to provide factual information. - While the model can output text, text rendered may be inaccurate or subject to distortion. - As a statistical model, this checkpoint may represent or amplify biases observed in the training data. - The model may fail to generate output that matches the prompts.…

Open weights apache-2.0 diffusion-single-file
View model

A LoRA adapter for the Qwen-Image-Edit-2511 diffusion pipeline, enabling NSFW content generation and editing capabilities. The easiest way to use this LoRA is through the ScottzillaSystems Image Editor — select " MCNL-NSFW-v1" from the adapter dropdown. Use these concepts in your prompts to activate specific capabilities: nsfw nipples vagina penis missionary cowgirlout reversecowgirlpov blowjob cumonface creamp1e l1ck

Open weights openrail++ diffusers
View model

FLUX.2 [klein] 4B Base is a 4 billion parameter rectified flow transformer capable of generating images from text descriptions and supports multi-reference editing capabilities. For more information, please read our blog post. This repository holds an FP8 version of FLUX.2 [klein] 4B Base. The main repository of this model (full BF16 weights) can be found here. Limitations - This model is not intended or able to provide factual information. - While the model can output text, text rendered may be inaccurate or subject to distortion. - As a statistical model, this checkpoint may represent or amplify biases observed in the training data. - The model may fail to generate output that matches the…

Open weights apache-2.0 diffusion-single-file
View model

Model · Image to image

safetensored-upscalers

Ashtaka

compact is quick and decent i guess - dat is very slow but great quality (dark go to) - drct similar quality to dat but slightly faster - esrgan is old and mediocre - nafnet is an image restoration architecture - omnisr is quick and good quality (nub go to) - rcan is old but good - real-esrgan is esrgan but real (mediocre) - span is quick and good quality (my go to) I do not own any models in this repo. These are just converted pickles to safetensors. The script used is modified from https://github.com/OpenModelDB/model-hub/blob/main/tools/convert-pth.py with Gemini 2.5 Pro.

Open weights mit
View model

Model · Image to image

stopmotion-consistency-v1-lora

SSH

Built with Qwen. An experimental next-frame image-editing adapter fine-tuned from the original BF16 Qwen-Image-2.1. The adapter was trained on 112 curated stop-motion transitions using an RTX PRO 6000. Quality is limited. Matched tests show improved static-area retention, but motion magnitude, direction and shape consistency remain unreliable. This is a small research pilot, not a proven solution for long, consistent animation. The adapter is stopmotion-consistency-v1.safetensors, 67,141,056 bytes. It contains only attention LoRA updates, not the base model. SHA-256: d68479ed2c246bc7ebea8a62019b10da903772d30854385302b2d66de99df51a. Use DiffSynth-Studio at commit…

Open weights other
View model