whatmodel.sh / image / FLUX.2 [klein] 9B
FLUX.2 [klein] 9B
↗ share
+ contribute
FLUX.2 [klein] 9B is an open-weights image model from Black Forest Labs, released January 2026. Its published size is 9B parameters. Licence: flux-dev-nc.
Score within the LMArena text-to-image comparison set only.
$ how it compares
The image models ranked either side of FLUX.2 [klein] 9B, out of 197 with a score, then more from Black Forest Labs.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 80 |  Qwen-Image-Edit | 55±2 | api $0.030/image +local | ||
| 81 | FLUX.2 [klein] Base 9B | 55±4 | comparison only | ||
| 82 | FLUX.2 [klein] 9B | 55±2 | api $0.015/image +local | ||
| 83 | FLUX.1 Kontext [max] | 55±2 | api $0.080/image | ||
| 84 |  Reve Edit Fast | 55±4 | comparison only | ||
| more from Black Forest Labs | |||||
| 32 | FLUX.2 [max] | 64±1 | api $0.070/image | ||
| 36 | FLUX.2 [pro] | 63±2 | api $0.030/image | ||
| 37 | FLUX.2 [flex] | 62±2 | api $0.050/image | ||
$ what it's good at
text to image▂▅▅▃▄▄█▃▁▂▁▁ #61 of 126 · 67% of its score · 3 tests
Trails LongCat-Image, LLaDA-Image-Turbo, FLUX.1 Kontext [max] here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| LMArena text-to-image | 1070.3 ✎ | ▂▂▃▄▇█▄▂▄▂·▂ beats 19/75 | 41% |
| QC-T2I-Bench (English, HCQ total) | 87.1% ✎ | ▄····▄█▄▄█▄█ beats 6/10 | 15% |
| UltraText Bench (Composite, bilingual dense text rendering) | 22.9% ✎ | ▁▂▃█▃▁▂▃·▁·▁ beats 3/22 | 11% |
image to image / editing▁▁·▃▂▃█▃▃▄▁▁ #34 of 62 * · 22% of its score · 2 tests
Beats FLUX.2 [flex], Imagen 4, GPT Image 1 Mini here, though they rank higher overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Arena image edit | 1224.7 ✎ | ▂▁▁▂▆▇█▄▆▅▁▂ beats 15/53 | 16% |
| GEditBench v2 (overall Elo) | 1039.0 ✎ | ▃▂····▂█▂▃▂▃ beats 11/14 | 6.2% |
other tests2 not in a task · 10% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Rapidata Benchmark.ai image benchmark (overall Elo) | 925.3 ✎ | ▁▁▁▄▂█▂▁▁▂▁▁ beats 6/34 | 6.4% |
| PaintBench precise editing (mIoU, 20-task average) | 3.5% ✎ | ▃·····█▅▃·▃▅ beats 1/9 | 3.1% |
$ where to run it
Through an API
api · Cloudflare · $0.015/image · developers.cloudflare.com ↗ · price sourceCloudflare Workers AI; USD0.015 for first MP (explicitly 1024×1024 in source); no input-image MP. Workers plan fee excluded; published USD equivalent is rounded, service bills Neurons. Product: Workers AI text-to-image. API model: @cf/black-forest-labs/flux-2-klein-9b. Settings: width=1024, height=1024, input images=0. Price basis: derived fixed-setting output cost using published rounded USD rates. Additional/excluded charges: Workers Paid plan base fee, input images and usage beyond this setting. Currency: USD.
Other ways to run it
version
- Run it as published · needs ~13.7 GB · from github.com/leejet
sd-cli --diffusion-model models/flux-2-klein-9b-Q4_0.gguf --vae models/flux2_ae.safetensors --llm models/Qwen3-8B-Q4_K_M.gguf -p "a lovely cat" --cfg-scale 1.0 --steps 4 -v --offload-to-cpu --diffusion-fa
Official weights and documentation: huggingface.co/black-forest-labs ↗
$ how the score was worked out · LMArena leaderboard dataset 57%, QC-T2I-Bench: Scalable Questi… 15%, UltraText Bench: A Comprehens… 11% and 3 more
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json