AI models
DiT-XL/2
Training compute
6×10²⁰ FLOP
Parameters
675M
Published
Mar 2, 2023
DiT-XL/2 is an AI model developed by New York University (NYU) and University of California (UC) Berkeley (United States), first published in March 2023. It works in the image generation domain, on tasks such as image generation.
Training it took an estimated 6×10²⁰ FLOP of compute (estimation method: hardware,other). The model has 675,000,000 parameters. Training ran on Google TPU v3. The compute alone is estimated at $111K in 2023 dollars.
Access: Open weights (non-commercial). Its weights are openly available. It is built on top of Stable Diffusion (LDM-KL-8-G). The reference paper has 5,753 citations. Epoch AI rates the confidence of this record as confident.
Full record
- Organization
- New York University (NYU), University of California (UC) Berkeley
- Country of organization
- United States
- Domain
- Image generation
- Task
- Image generation
- Training compute
- 6×10²⁰ FLOP
- Compute estimation method
- Hardware, Other
- Parameters
- 675,000,000
- Training hardware
- Google TPU v3
- Training cost (2023 USD)
- $111K
- Model accessibility
- Open weights (non-commercial)
- Open weights
- Yes
- Base model
- Stable Diffusion (LDM-KL-8-G)
- Citations
- 5,753
- Epoch confidence
- Confident
SourceEpoch AI, 'AI Models'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/data/ai-models. Licensed under CC BY 4.0.
← All ai models