Lumina-Image 2.0
2.6B open DiT image model that punches above its weight.
About
Lumina-Image 2.0, published March 27, 2025 as arXiv 2503.21758, is the Shanghai AI Lab's efficiency play in text-to-image: a 2.6-billion-parameter Unified Next-DiT that treats text and image tokens as one joint sequence, paired with a UniCap captioning system built specifically for generation training. The paper reports strong results on academic benchmarks and public text-to-image arenas despite the small size, crediting multi-stage progressive training and inference acceleration that doesn't cost quality. Code, training details, and weights are all public (Alpha-VLLM/Lumina-Image-2.0), and it runs through diffusers. This is a research artifact first: excellent if you want an open, efficient base to build on, not a polished product with an API and support.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Paper-reported: strong results on academic T2I benchmarks and public text-to-image arenas at 2.6B params (provider-reported)
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 7/10
Scored 7/10 (verified tier, metrics-v1). Evidence: Paper published Mar 27, 2025 (arXiv 2503.21758); 2.6B Unified Next-DiT from Shanghai AI Lab.; Open weights, code, and training details public (Alpha-VLLM/Lumina-Image-2.0).; Paper-reported strong results on academic benchmarks and public arenas at small size.; Limitations: research artifact — no hosted API or product polish..