Frederic Boesel
Frederic Boesel is a researcher whose work connects generative AI with synthetic biological data, cloud anomaly detection and high-resolution image synthesis. He studied mathematics at the University of Freiburg and data science at ETH Zurich, then worked as a researcher at IBM and Stability AI before joining Black Forest Labs. In 2024, he coauthored Latent Adversarial Diffusion Distillation, which produced SD3-Turbo with four unguided sampling steps, and a rectified-flow image-generation paper that won the ICML Best Paper award. In 2025, he coauthored FLUX.1 Kontext, a model for generating and editing images from text and image inputs. He is a founding member of Black Forest Labs, where he works on generative models such as the FLUX family.
- Now at
- Black Forest Labs
Timeline
-
2015–2019
Mathematics at Freiburg
He earns a bachelor’s degree in mathematics at the University of Freiburg in 2019.
-
2019–2023
ETH Zurich and IBM
From 2019 to 2020, he interns at BASF Digital Farming; in 2020, he works there as a data science working student. He studies data science at ETH Zurich from 2020 to 2023 and interns at IBM Research’s Thomas J. Watson Research Center from 2022 to 2023.
-
2020–2021
Generative models for omics
Between 2020 and 2021, he studies deep generative models for representing joint distributions in biological data, including small-sample omics datasets.
-
2023
Cloud anomaly detection
In 2023, he coauthors “Multivariate Anomaly Detection with Domain Clustering.” The paper introduces MADDoC, an unsupervised transfer learning framework for anomaly detection on multivariate time-series data across cloud infrastructure components.
-
2024
Diffusion distillation at four steps
In 2024, he coauthors Latent Adversarial Diffusion Distillation, a method applied to Stable Diffusion 3 to create SD3-Turbo with four unguided sampling steps.
-
2024
An ICML Best Paper
He coauthors “Scaling Rectified Flow Transformers for High-Resolution Image Synthesis,” which wins an ICML 2024 Best Paper award.
-
2025
FLUX.1 Kontext
In 2025, he coauthors FLUX.1 Kontext, a flow-matching model that unifies image generation and editing through text and image inputs.
-
Since 2024
At Black Forest Labs
He joins Black Forest Labs as a founding member in 2024 and works on generative models such as the FLUX family. Before joining the lab, he works as a researcher at Stability AI and IBM.
Career
- Since 2024Black Forest LabsFounding Member
- Until 2024Stability AIResearcher
- 2022–2023IBM ResearchIntern
- 2019–2020BASFIntern, Data science working student
Education
- 2023ETH Zurich
- 2019University of FreiburgBachelor's degree in mathematics
Sources and credits
Photo: We have no published photo of Frederic Boesel yet. Send us a free-licensed photo.
- Frederic Boesel (Black Forest Labs) — Lacuna
- Multivariate Anomaly Detection with Domain Clustering for SoCC 2023 - IBM Research
- Fast High-Resolution Image Synthesis with Latent Adversarial Diffusion Distillation
- ICML 2024 Awards
- FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Space
- Frederic Boesel × Forum UnternehmerTUM

