self-hosted/ai
§01·model · /models

Krea 2

imageactiveKrea 2 Community

Krea AI's first from-scratch, aesthetic-first foundation text-to-image model (released 2026-06-23). ~13B-parameter DiT with flow-matching, paired with a Qwen3-VL text encoder and the Qwen-Image VAE. Two open-weight variants under the Krea 2 Community License: Raw (CFG, full quality; 24.76 GiB BF16 transformer) and Turbo (distilled, ~8-step, ~2 s/image). Runs in ComfyUI 0.25.0+ natively with no custom nodes. The consumer entry point is the FP8 Turbo build (~12 GiB), which fits 16 GB+ GPUs. Commercial use is free for companies under $1M trailing-12-month revenue.

§02·GPUs that run this model
22 total
GPUVRAMSeriesBest speedMin VRAMWorksBenchmarksRecipe
RTX 309024GB3014.75s/img18.75GB1recipecheck ↗
Apple M2 Max64GBapple~0recipecheck ↗
Apple M3 Max48GBapple~0recipecheck ↗
Apple M4 Max48GBapple~0recipecheck ↗
RTX 306012GB30~0recipecheck ↗
RTX 3080 Ti12GB30~0recipecheck ↗
RTX 3090 Ti24GB30~0recipecheck ↗
RTX 4060 Ti 16GB16GB40~0recipecheck ↗
RTX 407012GB40~0recipecheck ↗
RTX 4070 Super12GB40~0recipecheck ↗
RTX 4070 Ti12GB40~0recipecheck ↗
RTX 4070 Ti Super16GB40~0recipecheck ↗
RTX 408016GB40~0recipecheck ↗
RTX 4080 Super16GB40~0recipecheck ↗
RTX 409024GB40~0recipecheck ↗
RTX 5060 Ti16GB50~0recipecheck ↗
RTX 507012GB50~0recipecheck ↗
RTX 5070 Ti16GB50~0recipecheck ↗
RTX 508016GB50~0recipecheck ↗
RTX 509032GB50~0recipecheck ↗
RX 7800 XT16GBamd~0recipecheck ↗
RX 7900 XTX24GBamd~0recipecheck ↗

benchmarked·~ runs via recipe (not benchmarked)· untested·doesn't fit