# FLUX.2-small-decoder by Black Forest Labs: Open-Weight Model Source: https://savrn.com/models/flux-2-small-decoder Markdown alternate of the page above; the site index is https://savrn.com/llms.txt --- ## Runs On What it takes to serve FLUX.2-small-decoder (62M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them. | Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits | | --- | --- | --- | --- | --- | --- | | 16-bit | 0.1 GB | 0.1 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 8-bit | 0.1 GB | 0.1 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | | 4-bit | 0.0 GB | 0.0 GB | 1x [MI300X](https://savrn.com/ai-index/pricing/gpus/mi300x) (192 GB) Vultr | $1.85 | [1x H100](https://savrn.com/ai-index/pricing/gpus/h100) $1.99 · [1x MI325X](https://savrn.com/ai-index/pricing/gpus/mi325x) $2.00 | Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the [SAVRN Index](https://savrn.com/ai-index/pricing/gpus), read Oct 24, 2026. [FLUX.2-small-decoder on every accelerator the SAVRN Index prices, at every precision](https://savrn.com/models/flux-2-small-decoder/gpus) ## Model Card By Black Forest Labs, published under apache-2.0, revision a3efc24f613e. FLUX.2 Small Decoder is a distilled VAE decoder that serves as a drop-in replacement for the standard FLUX.2 decoder. It delivers faster decoding and lower VRAM usage with minimal to zero quality loss. The encoder remains unchanged. ### Key Features 1. ~1.4x faster decoding compared to the full decoder. 2. ~1.4x less VRAM at decode time, enabling higher resolutions without running out of memory. 3. ~28M decoder parameters (vs ~50M in the full decoder) thanks to narrower channel widths ([96, 192, 384, 384] vs [128, 256, 512, 512]). 4. Minimal quality loss — images are almost identical. 5. Available under the Apache 2.0 license. Compatible with all open FLUX.2 models: - [FLUX.2-klein-4B](https://savrn.com/models/flux-2-klein-4b) - [FLUX.2-klein-9B](https://huggingface.co/black-forest-labs/FLUX.2-klein-9B) - [FLUX.2-klein-9b-kv](https://huggingface.co/black-forest-labs/FLUX.2-klein-9b-kv) - [FLUX.2-dev](https://huggingface.co/black-forest-labs/FLUX.2-dev) ### Comparison | Full Decoder | Small Decoder | | --- | --- | ### Usage ``` pip install git+https://github.com/huggingface/diffusers.git ``` [Read the full model card (402 words)](https://savrn.com/models/flux-2-small-decoder/card) ## Identity and Version Repository black-forest-labs/FLUX.2-small-decoder Publisher Black Forest Labs Task Image to image Modality Image Library diffusers Parameters 62M parameters Languages en Revision a3efc24f613ef42d9428af62fdbd6f5fd8856c4a First published 2026-04-06 Last updated 2026-04-07 ## Files and Weights 10 files, 619.4 MB in total. The weights are 3 files totalling 610.8 MB in safetensors. Weights3 files · 610.8 MB Configuration1 file · 842 B Documentation1 file · 3.9 KB Other4 files · 8.5 MB Repository1 file · 1.7 KB Every file | File | Type | Size | SHA-256 | | --- | --- | --- | --- | | diffusion_pytorch_model.safetensors | Weights | 249.5 MB | d8d52ba03647 | | full_encoder_small_decoder.safetensors | Weights | 249.5 MB | ea4273f02d1f | | small_decoder.safetensors | Weights | 111.8 MB | ad7833477475 | | config.json | Configuration | 842 B | — | | README.md | Documentation | 3.9 KB | — | | compare_full_decoder.png | Other | 1.2 MB | 188624cc1d06 | | compare_small_decoder.png | Other | 1.3 MB | 3ceccc1bea55 | | comparison_panel.jpeg | Other | 3.5 MB | ea01b25d80d3 | | editing.jpg | Other | 2.5 MB | 2912ca5a7cb9 | | .gitattributes | Repository | 1.7 KB | — | ## License and Download License apache-2.0 Access Open weights, no gate Download size 610.8 MB [Download from Black Forest Labs](https://huggingface.co/black-forest-labs/FLUX.2-small-decoder) Released by Black Forest Labs through its official repository on Hugging Face. [Read the license](https://www.apache.org/licenses/LICENSE-2.0). ## Memory Requirements | Precision | Weights in memory | | --- | --- | | As published | 610.8 MB | | 16-bit | 0.1 GB | | 8-bit | 0.1 GB | | 4-bit | 0.0 GB | Weights only, from the published parameter count; the key-value cache and runtime add to this. ## Questions About FLUX.2-small-decoder ### How much GPU memory does FLUX.2-small-decoder need? About 0.1 GB at 16-bit and 0 GB at 4-bit: the weights (62M parameters) plus a working margin. A long context needs more. ### What is the cheapest GPU to run FLUX.2-small-decoder on? At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists. ### Can I use FLUX.2-small-decoder commercially? Yes. FLUX.2-small-decoder 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 ### [photo-to-sketch-unet](https://savrn.com/models/photo-to-sketch-unet) [WILLIAMS OPOKU ANTWI](https://savrn.com/model-publishers/willie999) This model is a custom PyTorch U-Net architecture (SketchUNet) trained to convert RGB photos into target outline sketches using the rhfeiyang/photo-sketch-pair-500 dataset. - Combined Sketch Loss (L1 + Sobel Edge Loss): 0.14555 - Mean Absolute Error (L1 / MAE): 0.04962 This model converts an RGB image into a grayscale sketch using a PyTorch U-Net architecture. The model checkpoint is available as: Once the model architecture is defined, inference can be reduced to Open weights mit 31M parameters [View model](https://savrn.com/models/photo-to-sketch-unet) Model · Image to image ### [sol-image-edit](https://savrn.com/models/sol-image-edit) [Meana](https://savrn.com/model-publishers/itriedcoding) Open weights mit 11M parameters pytorch [View model](https://savrn.com/models/sol-image-edit) Model · Image to image ### [mini-unet-colorizer](https://savrn.com/models/mini-unet-colorizer) [Joe Bloggs](https://savrn.com/model-publishers/user-2468) Status: evaluated app-testing candidate. Some broad color patches and incorrect object hues remain. Predicted colors are not evidence of original historical colors. This checkpoint has 3,968,892 learned parameters and 236 fixed color bins. It starts from the audited bin-mapping repair of main commit 6c47ea40724d8fcd67d4f36ce837dc1cb5b1b2a8 and changes all 65 learned parameter tensors. The color vocabulary is unchanged. The selected weights are update 748 of a completed 1,122-update BF16 L4 run using a 17,325-photo mixed training pool (11,943 Imagenette plus 5,382 COCO), batch 32, initial learning rate 1e-5, weighted classification loss and frozen BatchNorm running statistics. Selection… Open weights apache-2.0 4M parameters [View model](https://savrn.com/models/mini-unet-colorizer) Model · Image to image ### [Qens](https://savrn.com/models/qens) [Android](https://savrn.com/model-publishers/user-android) pipelinetag: image-to-image Open weights 707,011 parameters [View model](https://savrn.com/models/qens) Model · Image to image ### [MadStickArt](https://savrn.com/models/madstickart) [Danger Room](https://savrn.com/model-publishers/danger-room) MadStickArt is a trained, 61,296-parameter neural raster decoder for stick-figure storyboard scenes. It reconstructs 956 × 400 grayscale frames from four explicit layout masks: actors, setting, props, and motion. Its output contract is 2.39:1. Try MadStickArt online → — the browser trial is deployed separately. Its deployed model/version and renderer are recorded in the Space's model-provenance.json. The renderer included with this release supports 186 settings and 320 props; the deployed trial has its own model and renderer pins and is updated through a separate reviewed stage. Browser suggestions use transparent rules, while Strands decisions run in the downloadable Python pipeline. The… Open weights apache-2.0 61,296 parameters pytorch [View model](https://savrn.com/models/madstickart) Model · Image to image ### [MadStickArt-v0.8.0](https://savrn.com/models/madstickart-v0-8-0) [Danger Room](https://savrn.com/model-publishers/danger-room) MadStickArt is a trained, 61,296-parameter neural raster decoder for stick-figure storyboard scenes. It reconstructs 956 × 400 grayscale frames from four explicit layout masks: actors, setting, props, and motion. Its output contract is 2.39:1. Try MadStickArt online → — the browser trial is deployed separately. Its deployed model/version and renderer are recorded in the Space's model-provenance.json. The renderer included with this release supports 186 settings and 320 props; the deployed trial has its own model and renderer pins and is updated through a separate reviewed stage. Browser suggestions use transparent rules, while Strands decisions run in the downloadable Python pipeline. The… Open weights apache-2.0 61,296 parameters pytorch [View model](https://savrn.com/models/madstickart-v0-8-0) ## Black Forest Labs [All models and datasets](https://savrn.com/model-publishers/black-forest-labs) ## Versions - [a3efc24f613e](https://savrn.com/models/flux-2-small-decoder/versions/a3efc24f613e) · current 2026-09-18 ## Explore More - [All image to image models](https://savrn.com/models/tasks/image-to-image) - [All models under apache-2.0](https://savrn.com/models/licenses/apache-2-0) - [Model comparisons](https://savrn.com/models/comparisons) - [The model directory](https://savrn.com/models) - [Open model prices by host](https://savrn.com/ai-index/pricing/open-models) ## Source - Repository metadata, read 2026-10-20. - [Hugging Face record](https://huggingface.co/black-forest-labs/FLUX.2-small-decoder) - [How the hub is built](https://savrn.com/model-hub/methodology)