{"id":1907,"date":"2026-04-20T09:13:37","date_gmt":"2026-04-20T09:13:37","guid":{"rendered":"https:\/\/wiro.ai\/blog\/?p=1907"},"modified":"2026-03-21T14:25:37","modified_gmt":"2026-03-21T14:25:37","slug":"flux-2-klein-base-4b-5-real-image-tests","status":"publish","type":"post","link":"https:\/\/wiro.ai\/blog\/flux-2-klein-base-4b-5-real-image-tests\/","title":{"rendered":"FLUX.2 Klein Base 4B: 5 Real Image Tests"},"content":{"rendered":"

FLUX.2 Klein Base 4B: what it is<\/h2>\n

FLUX.2 Klein Base 4B is an open-weight image model from Black Forest Labs. It targets fast image generation and editing in a small footprint. The \u201cBase\u201d variant matters because it stays undistilled. That usually means more output diversity and more room for fine-tuning.<\/p>\n

This post runs five small tests. Four are text-to-image. One is image-to-image editing from a single reference photo.<\/p>\n

Test setup<\/h2>\n\n\n\n\n\n\n\n
Setting<\/th>\nValue<\/th>\n<\/tr>\n
Model<\/td>\nblack-forest-labs\/flux-2-klein-base-4b<\/a><\/td>\n<\/tr>\n
Resolution<\/td>\n1024\u00d71024<\/td>\n<\/tr>\n
Steps<\/td>\n30<\/td>\n<\/tr>\n
Guidance scale<\/td>\n4.0<\/td>\n<\/tr>\n
Seeds<\/td>\n101\u2013105 (one per test)<\/td>\n<\/tr>\n<\/table>\n

Text-to-image tests (4 prompts)<\/h2>\n

1) Product photo realism<\/h3>\n

This prompt checks materials, lighting, and small details like condensation.<\/p>\n

\n \"AI
Prompt: Commercial product photo of a matte black travel mug on dark slate, condensation beads on metal, softbox key light from left, subtle rim light, shallow depth of field, photorealistic, 50mm lens look<\/figcaption><\/figure>\n

2) Night street scene + readable neon text<\/h3>\n

This prompt pushes signage clarity in a busy scene with reflections.<\/p>\n

\n \"AI
Prompt: Candid night street photo in Seoul after rain, neon shop sign must read NOODLES in clear block letters, wet asphalt reflections, people with umbrellas in the distance, 35mm film grain, high contrast, photorealistic<\/figcaption><\/figure>\n

3) Macro detail<\/h3>\n

This prompt checks fine texture, focus control, and natural lighting.<\/p>\n

\n \"AI
Prompt: Extreme macro photo of a honey bee landing on a lavender flower, visible pollen grains on fuzzy legs, translucent wing veins, creamy green bokeh background, razor sharp focus on eyes, natural sunlight, 100mm macro lens look<\/figcaption><\/figure>\n

4) Simple typography on a real object<\/h3>\n

Instead of a full poster, this uses a street sign style layout. It still tests letter shapes and spacing.<\/p>\n

\n \"AI
Prompt: Photo of a vintage metal street sign mounted on a red brick wall, crisp bold typography, text must read BAY STREET on the top line and SAN FRANCISCO on the bottom line, shallow depth of field, natural daylight, photorealistic<\/figcaption><\/figure>\n

Image-to-image edit test (single reference)<\/h2>\n

This test starts from a reference dog photo. The prompt asks for a 1970s film still look while keeping the pose.<\/p>\n

\n \"Input
Input image used for the edit test<\/figcaption><\/figure>\n
\n \"Edited
Prompt: Turn this into a vintage 1970s film still, warm tungsten lighting, subtle grain, shallow depth of field, keep the dog pose and fur detail realistic, cinematic color grade<\/figcaption><\/figure>\n

What looked strong (and what didn\u2019t)<\/h2>\n