Halcyon early-fusion image models match late-fusion with 275x less pretraining compute, open-sourced
davidwromero · x · 2026-10-16
Radical Numerics introduced Halcyon, a family of early-fusion base models trained from scratch, asking whether text-to-image generation really needs language pretraining.
- Halcyon outperforms strong late-fusion baselines while using up to 275× less pretraining compute
- Models and training recipes are open-sourced
- Part of the company's research program on the science of multimodal pretraining and scaling
Related event: Halcyon: Text-to-Image Base Models That Cut Pre-training Compute 275x(3 posts)→
More from Multimodal
- Krea 2 + LoRA still beats Flux 3 Klein in pixel-art style test — Angaisb_ · 2026-10-16
- World Labs starts rolling out Atlas beta with pixel-perfect camera path control — theworldlabs · 2026-10-16
- Krea 3 image editing model lands on fal: high-quality edits in under 6 seconds — gorkem · 2026-10-16
- Tripo P2 vs Meshy 7.1: benchmarking today's best image-to-3D models for game assets — noahsolomon · 2026-10-16
- BytePlus-backed AI Film Award offers $1M grand prize for films 70% made with Seedance — DavidmComfort · 2026-10-16
- Halcyon: from-scratch early-fusion T2I models match late fusion with 275x less pretraining compute — StanfordAILab · 2026-10-16