# 1.58-bit FLUX

arXiv: https://arxiv.org/abs/2412.18653 · HTML: https://mlsanity.com/paper/2412.18653/

## Key facts

- Cited by 31 papers in the map (data through 14 Oct 2026).
- Builds on 31 of its 63 references that are in the map.
- First posted on arXiv 24 Dec 2024.

## Builds on (most cited first, 30 of 31)

- [SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis](https://mlsanity.com/paper/2307.01952/) (2023)
- [Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding](https://mlsanity.com/paper/1510.00149/) (2015)
- [GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers](https://mlsanity.com/paper/2210.17323/) (2022)
- [Scaling Autoregressive Models for Content-Rich Text-to-Image Generation](https://mlsanity.com/paper/2206.10789/) (2022)
- [PixArt-α: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesis](https://mlsanity.com/paper/2310.00426/) (2023)
- [Learned Step Size Quantization](https://mlsanity.com/paper/1902.08153/) (2019)
- [A White Paper on Neural Network Quantization](https://mlsanity.com/paper/2106.08295/) (2021)
- [OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models](https://mlsanity.com/paper/2308.13137/) (2023)
- [LLM-QAT: Data-Free Quantization Aware Training for Large Language Models](https://mlsanity.com/paper/2305.17888/) (2023)
- [The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits](https://mlsanity.com/paper/2402.17764/) (2024)
- [SqueezeLLM: Dense-and-Sparse Quantization](https://mlsanity.com/paper/2306.07629/) (2023)
- [Playground v2.5: Three Insights towards Enhancing Aesthetic Quality in Text-to-Image Generation](https://mlsanity.com/paper/2402.17245/) (2024)
- [QServe: W4A8KV4 Quantization and System Co-design for Efficient LLM Serving](https://mlsanity.com/paper/2405.04532/) (2024)
- [BitNet: Scaling 1-bit Transformers for Large Language Models](https://mlsanity.com/paper/2310.11453/) (2023)
- [SVDQuant: Absorbing Outliers by Low-Rank Components for 4-Bit Diffusion Models](https://mlsanity.com/paper/2411.05007/) (2024)
- [EfficientQAT: Efficient Quantization-Aware Training for Large Language Models](https://mlsanity.com/paper/2407.11062/) (2024)
- [Playground v3: Improving Text-to-Image Alignment with Deep-Fusion Large Language Models](https://mlsanity.com/paper/2409.10695/) (2024)
- [ViDiT-Q: Efficient and Accurate Quantization of Diffusion Transformers for Image and Video Generation](https://mlsanity.com/paper/2406.02540/) (2024)
- [Q-DiT: Accurate Post-Training Quantization for Diffusion Transformers](https://mlsanity.com/paper/2406.17343/) (2024)
- [PTQ4DiT: Post-training Quantization for Diffusion Transformers](https://mlsanity.com/paper/2405.16005/) (2024)
- [EfficientDM: Efficient Quantization-Aware Fine-Tuning of Low-Bit Diffusion Models](https://mlsanity.com/paper/2310.03270/) (2023)
- [MixDQ: Memory-Efficient Few-Step Text-to-Image Diffusion Models with Metric-Decoupled Mixed Precision Quantization](https://mlsanity.com/paper/2405.17873/) (2024)
- [BitsFusion: 1.99 bits Weight Quantization of Diffusion Model](https://mlsanity.com/paper/2406.04333/) (2024)
- [EDA-DM: Enhanced Distribution Alignment for Post-Training Quantization of Diffusion Models](https://mlsanity.com/paper/2401.04585/) (2024)
- [HQ-DiT: Efficient Diffusion Transformer with FP4 Hybrid Quantization](https://mlsanity.com/paper/2405.19751/) (2024)
- [Efficient Quantization Strategies for Latent Diffusion Models](https://mlsanity.com/paper/2312.05431/) (2023)
- [Timestep-Aware Correction for Quantized Diffusion Models](https://mlsanity.com/paper/2407.03917/) (2024)
- [VQ4DiT: Efficient Post-Training Vector Quantization for Diffusion Transformers](https://mlsanity.com/paper/2408.17131/) (2024)
- [PrefixQuant: Eliminating Outliers by Prefixed Tokens for Large Language Models Quantization](https://mlsanity.com/paper/2410.05265/) (2024)
- [1-bit AI Infra: Part 1.1, Fast and Lossless BitNet b1.58 Inference on CPUs](https://mlsanity.com/paper/2410.16144/) (2024)

## Built on by (most cited first, 30 of 31)

- [Beyond Next-Token: Next-X Prediction for Autoregressive Visual Generation](https://mlsanity.com/paper/2502.20388/) (2025)
- [Democratizing Text-to-Image Masked Generative Models with Compact Text-Aware One-Dimensional Tokens](https://mlsanity.com/paper/2501.07730/) (2025)
- [ParetoQ: Improving Scaling Laws in Extremely Low-bit LLM Quantization](https://mlsanity.com/paper/2502.02631/) (2025)
- [Kling-Foley: Multimodal Diffusion Transformer for High-Quality Video-to-Audio Generation](https://mlsanity.com/paper/2506.19774/) (2025)
- [Training-Free Efficient Video Generation via Dynamic Token Carving](https://mlsanity.com/paper/2505.16864/) (2025)
- [FlowTok: Flowing Seamlessly Across Text and Image Tokens](https://mlsanity.com/paper/2503.10772/) (2025)
- [From One to More: Contextual Part Latents for 3D Generation](https://mlsanity.com/paper/2507.08772/) (2025)
- [Skywork UniPic: Unified Autoregressive Modeling for Visual Understanding and Generation](https://mlsanity.com/paper/2508.03320/) (2025)
- [Plug-and-Play 1.x-Bit KV Cache Quantization for Video Large Language Models](https://mlsanity.com/paper/2503.16257/) (2025)
- [Grouping First, Attending Smartly: Training-Free Acceleration for Diffusion Transformers](https://mlsanity.com/paper/2505.14687/) (2025)
- [OmniPSD: Layered PSD Generation with Diffusion Transformer](https://mlsanity.com/paper/2512.09247/) (2025)
- [Text-to-Image Diffusion Models Cannot Count, and Prompt Refinement Cannot Help](https://mlsanity.com/paper/2503.06884/) (2025)
- [SeaCache: Spectral-Evolution-Aware Cache for Accelerating Diffusion Models](https://mlsanity.com/paper/2602.18993/) (2026)
- [A Frame is Worth One Token: Efficient Generative World Modeling with Delta Tokens](https://mlsanity.com/paper/2604.04913/) (2026)
- [Deeply Supervised Flow-Based Generative Models](https://mlsanity.com/paper/2503.14494/) (2025)
- [HieraTok: Multi-Scale Visual Tokenizer Improves Image Reconstruction and Generation](https://mlsanity.com/paper/2509.23736/) (2025)
- [Diffusion Classifiers Understand Compositionality, but Conditions Apply](https://mlsanity.com/paper/2505.17955/) (2025)
- [Adaptive Divergence Regularized Policy Optimization for Fine-tuning Generative Models](https://mlsanity.com/paper/2510.18053/) (2025)
- [Frequency-Aware Flow Matching for High-Quality Image Generation](https://mlsanity.com/paper/2604.15521/) (2026)
- [Trans-Adapter: A Plug-and-Play Framework for Transparent Image Inpainting](https://mlsanity.com/paper/2508.01098/) (2025)
- [DFVEdit: Conditional Delta Flow Vector for Zero-shot Video Editing](https://mlsanity.com/paper/2506.20967/) (2025)
- [Investigation of Accuracy and Bias in Face Recognition Trained with Synthetic Data](https://mlsanity.com/paper/2507.20782/) (2025)
- [DeCoT: Decomposing Complex Instructions for Enhanced Text-to-Image Generation with Large Language Models](https://mlsanity.com/paper/2508.12396/) (2025)
- [IRIS: Intrinsic Reward Image Synthesis](https://mlsanity.com/paper/2509.25562/) (2025)
- [WeMMU: Enhanced Bridging of Vision-Language Models and Diffusion Models via Noisy Query Tokens](https://mlsanity.com/paper/2512.02536/) (2025)
- [When Do Diffusion Models learn to Generate Multiple Objects?](https://mlsanity.com/paper/2605.00273/) (2026)
- [Ternarization of Vision Language Models for use on edge devices](https://mlsanity.com/paper/2504.06298/) (2025)
- [ID-Card Synthetic Generation: Toward a Simulated Bona fide Dataset](https://mlsanity.com/paper/2508.13078/) (2025)
- [Synergizing Understanding and Generation with Interleaved Analyzing-Drafting Thinking](https://mlsanity.com/paper/2602.21435/) (2026)
- [RT-Lynx: Putting GEMM Sparsity in the Right Place for Diffusion Models](https://mlsanity.com/paper/2605.26632/) (2026)

## Abstract (arXiv)

We present 1.58-bit FLUX, the first successful approach to quantizing the state-of-the-art text-to-image generation model, FLUX.1-dev, using 1.58-bit weights (i.e., values in {-1, 0, +1}) while maintaining comparable performance for generating 1024 x 1024 images. Notably, our quantization method operates without access to image data, relying solely on self-supervision from the FLUX.1-dev model. Additionally, we develop a custom kernel optimized for 1.58-bit operations, achieving a 7.7x reduction in model storage, a 5.1x reduction in inference memory, and improved inference latency. Extensive evaluations on the GenEval and T2I Compbench benchmarks demonstrate the effectiveness of 1.58-bit FLUX in maintaining generation quality while significantly enhancing computational efficiency.