Image Generation
Image-generation models turn a text prompt (or an existing image) into a new image — Midjourney, Stable Diffusion, DALL-E, and Flux are the category’s best-known names, spanning both closed commercial products and open-weight models builders can self-host. The competitive axes here are photorealism, prompt adherence, editing control, and increasingly licensing terms for commercial use. GROUNDING tracks new image-model releases and the techniques (diffusion architecture changes, distillation for speed) driving quality and cost improvements in this category.
At a glance
- 10 tracked models (1 hand-curated)
- Most recently updated: Muse Image (2026-07-08)
Most actively covered
- Muse Image
- GPT-Image-2
- Nano Banana 2 Lite
- Nano Banana Pro
- Nano Banana
- FLUX.1-dev
- DALL-E
- Gemini 3.1 Flash Image
- Stable Diffusion
- DiffusionGemma ★
FAQ
What is the Image Generation category?
Text-to-image and image-editing generative models.
How many models are in this category?
GROUNDING currently tracks 10 models in Image Generation, 1 of which have a hand-curated profile.
How current is this hub?
The most recently updated entry is Muse Image, last mentioned 2026-07-08; the radar refreshes hourly.