whatmodel.sh · find the best image models you can run

Best image editing models for a 32GB graphics card

A leaderboard of the best image editing models for a 32GB graphics card (with 32GB of system RAM). The best one that runs on it right now is Qwen-Image-2.1, followed by Qwen-Image-Edit-2509 and FLUX.2 [klein] 9B. Models are ranked on human votes and image editing tests. Models that run on it show the download that fits and the command to run it.

62 models · scores combine 15 benchmarks · updated how it works · find a leaderboard · benchmark your machine
task[all tasks][text to image][image to image / editing]

Scores use reviewed image to image / editing evaluations only. Compare scores within this task. Models without task measurements are excluded.

Models for 32GB RAM · 32GB graphics card → ~32GB fast · reset
+ more
card[RTX 5090][RTX 5000 Ada][RTX PRO 4500 Blackwell]

Detect your exact chip or graphics card:

curl -G whatmodel.sh/image -d max=0 --data-urlencode "mac=$(sysctl -n hw.memsize machdep.cpu.brand_string)"Mac

curl -G whatmodel.sh/image -d max=0 --data-urlencode "ram=$(awk '/MemT/{print $2}' /proc/meminfo)" --data-urlencode "nv=$(nvidia-smi --query-gpu=name,memory.total --format=csv,noheader 2>/dev/null)"Linux

curl.exe -G whatmodel.sh/image -d max=0 --data-urlencode "ram=$((gcim Win32_ComputerSystem).TotalPhysicalMemory)" --data-urlencode "nv=$(nvidia-smi --query-gpu=name,memory.total --format=csv,noheader 2>$null)"Windows

$ terminal

detecting your OS…

Run curl "whatmodel.sh/image?ram=32&gpu=card&vram=32"to get this list in your terminal (add &format=json for agents).

ram[8][16][24][32][48][64][96][128][192][256]GB
gpu[apple silicon][graphics card][none]
vram[1][2][4][8][12][16][24][32][48][64][80][96][128][160]GB
price[local][$0.002][$0.005][$0.01][$0.02][$0.05][$0.1][any]max per image

detecting your OS…

Score vs API price

#modeltask score↓where↓speed↓RAM↓get itsources
#modeltask score↓where↓speed↓RAM↓get itsources
1GPT Image 2.5 Sunburst OpenAI83±3*api $0.013/imagefal.ai ↗LMArena leaderboard dataset
2GPT Image 2.5 Flare OpenAI78±3*api $0.013/imagefal.ai ↗LMArena leaderboard dataset
3GPT Image 2 OpenAI76±3*api $0.053/imagefal.ai ↗LMArena leaderboard dataset
5Nano Banana 2.1 Google72±3*api $0.034/imageai.google.dev ↗LMArena leaderboard dataset
6Grok Imagine Image 2.0 SpaceXAI72±3*api $0.040/imagedocs.x.ai ↗LMArena leaderboard dataset
7Muse Image Meta69±3*api $0.010/imagedev.meta.ai ↗LMArena leaderboard dataset
9Nano Banana Pro (Gemini 3 Pro Image) Google68±2api $0.13/imageai.google.dev ↗Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward ModelingLMArena leaderboard datasetGEditBench v2: A Human-Aligned Benchmark for General Image EditingTECCI: Tricky Edits of Collected and Curated ImagesGEBench: Benchmarking Image Generation Models as GUI Environments
10Seedream 5.0 Pro ByteDance Seed68±3*api $0.050/imagedeepinfra.com ↗LMArena leaderboard dataset
11ChatGPT Image Latest (chatgpt-image-latest) OpenAI68±3*api $0.034/imagedevelopers.openai.com ↗LMArena leaderboard dataset
12Nano Banana 2 (Gemini 3.1 Flash Image) Google68±2api $0.067/imageai.google.dev ↗Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward ModelingLMArena leaderboard datasetTECCI: Tricky Edits of Collected and Curated ImagesMultiTextEdit: Benchmarking Cross-Lingual Degradation in Text-in-Image Editing (arXiv:2605.08163v2), Table 3
13Reve 2.1 Reve66±3*api $0.25/imagefal.ai ↗LMArena leaderboard dataset
14Grok Imagine Image Pro SpaceXAI (xAI)66±7*api $0.050/imagedocs.x.ai ↗TECCI: Tricky Edits of Collected and Curated Images
15Qwen-Image-2.1 · int8 convrot + Qwen3-VL 8B q4_k_m (sd.cpp) Alibaba66±3*local~16GBhow to run › LMArena leaderboard dataset
16Reve 2.0 Reve65±3*api $0.25/imagefal.ai ↗LMArena leaderboard dataset
18GPT Image 1.5 OpenAI63±2api $0.034/imagedevelopers.openai.com ↗LMArena leaderboard datasetGEditBench v2: A Human-Aligned Benchmark for General Image EditingGEBench: Benchmarking Image Generation Models as GUI EnvironmentsTECCI: Tricky Edits of Collected and Curated ImagesMultiTextEdit: Benchmarking Cross-Lingual Degradation in Text-in-Image Editing (arXiv:2605.08163v2), Table 3
19Uni-1.1 Max Luma AI62±3*api $0.10/imagelumalabs.ai ↗LMArena leaderboard dataset
20Grok Imagine Image SpaceXAI (xAI)61±3*api $0.020/imagedocs.x.ai ↗LMArena leaderboard dataset
21Uni-1.1 Luma AI60±3*api $0.040/imagelumalabs.ai ↗LMArena leaderboard dataset
22Nano Banana 2 Lite (Gemini 3.1 Flash-Lite Image) Google59±3*api $0.034/imageai.google.dev ↗LMArena leaderboard dataset
23Seedream 5.0 Lite ByteDance Seed59±3*api $0.035/imagedocs.byteplus.com ↗LMArena leaderboard datasetTECCI: Tricky Edits of Collected and Curated Images
24Qwen-Image-2.0-Pro (2026-06-22) Alibaba58±3*api $0.075/imagewww.alibabacloud.com ↗LMArena leaderboard dataset
25HunyuanImage 3.0-Instruct Tencent58±3*api $0.090/image~203GBfal.ai ↗LMArena leaderboard dataset
26Wan 2.7 Image Pro Alibaba58±3*api $0.075/imagewww.alibabacloud.com ↗LMArena leaderboard dataset
27Seedream 4.5 ByteDance58±2api $0.040/imagedocs.byteplus.com ↗Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward ModelingLMArena leaderboard datasetGEditBench v2: A Human-Aligned Benchmark for General Image EditingGEBench: Benchmarking Image Generation Models as GUI Environments
28Wan 2.7 Image Alibaba58±2*api $0.030/imagewww.alibabacloud.com ↗Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward ModelingLMArena leaderboard dataset
30FLUX.2 [max] Black Forest Labs54±3*api $0.070/imagedocs.bfl.ai ↗LMArena leaderboard dataset
32FLUX.2 [pro] Black Forest Labs53±3*api $0.030/imagedocs.bfl.ai ↗LMArena leaderboard datasetGEBench: Benchmarking Image Generation Models as GUI Environments
33Kling Image O1 Kuaishou (Kling AI)53±3*api $0.028/imagekling.ai ↗LMArena leaderboard dataset
34Nano Banana (Gemini 2.5 Flash Image) Google52±2api $0.039/imageai.google.dev ↗LMArena leaderboard datasetMeituan LongCat — LongCat-Image technical reportServImage: An Image Generation and Editing Benchmark from Real-world Commercial Imaging Services (arXiv:2604.24023v2)UnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched EditsGEBench: Benchmarking Image Generation Models as GUI EnvironmentsCREval: An Automated Interpretable Evaluation for Creative Image Manipulation under Complex Instructions (arXiv:2603.26174v1)+1
35Seedream 4.0 ByteDance Seed52±2api $0.040/imagedeepinfra.com ↗LMArena leaderboard datasetMeituan LongCat — LongCat-Image technical reportServImage: An Image Generation and Editing Benchmark from Real-world Commercial Imaging Services (arXiv:2604.24023v2)UnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched EditsGEBench: Benchmarking Image Generation Models as GUI EnvironmentsCREval: An Automated Interpretable Evaluation for Creative Image Manipulation under Complex Instructions (arXiv:2603.26174v1)
36Qwen-Image-Edit-2509 · gguf-q4_k_s + Qwen2.5-VL q8_0 (sd.cpp) Alibaba51±3local~26GBhow to run › Meituan LongCat — LongCat-Image technical reportGEditBench v2: A Human-Aligned Benchmark for General Image EditingBAAI — OmniGen2 technical report
37GPT Image 1 OpenAI51±2api $0.042/imagedevelopers.openai.com ↗Meituan LongCat — LongCat-Image technical reportLMArena leaderboard datasetServImage: An Image Generation and Editing Benchmark from Real-world Commercial Imaging Services (arXiv:2604.24023v2)UnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched EditsGEBench: Benchmarking Image Generation Models as GUI EnvironmentsCREval: An Automated Interpretable Evaluation for Creative Image Manipulation under Complex Instructions (arXiv:2603.26174v1)+1
38Qwen-Image-Edit-2511 Alibaba51±2api $0.030/image +local~37GBfal.ai ↗Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward ModelingLMArena leaderboard datasetGEditBench v2: A Human-Aligned Benchmark for General Image EditingMultiTextEdit: Benchmarking Cross-Lingual Degradation in Text-in-Image Editing (arXiv:2605.08163v2), Table 3
39FLUX.2 [klein] 9B · bf16 (cpu offload) Black Forest Labs51±3*local~22GBhow to run › LMArena leaderboard datasetGEditBench v2: A Human-Aligned Benchmark for General Image Editing
41LongCat-Image-Edit Meituan50±2api $0.15/image +local~36GBfal.ai ↗Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward ModelingMeituan LongCat — LongCat-Image technical reportGEditBench v2: A Human-Aligned Benchmark for General Image Editing
42Wan 2.6 Image Alibaba50±3*api $0.030/imagewww.alibabacloud.com ↗LMArena leaderboard dataset
43FLUX.2 [flex] Black Forest Labs50±3*api $0.050/imagedocs.bfl.ai ↗LMArena leaderboard dataset
46Imagen 4 Google49±5*api $0.040/imagecloud.google.com ↗ServImage: An Image Generation and Editing Benchmark from Real-world Commercial Imaging Services (arXiv:2604.24023v2)
47Qwen-Image-Edit · gguf-q4_k_m + Qwen2.5-VL q8_0 (sd.cpp) Alibaba49±2local~27GBhow to run › Meituan LongCat — LongCat-Image technical reportLMArena leaderboard datasetGEditBench v2: A Human-Aligned Benchmark for General Image EditingUnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched EditsGEBench: Benchmarking Image Generation Models as GUI EnvironmentsMultiTextEdit: Benchmarking Cross-Lingual Degradation in Text-in-Image Editing (arXiv:2605.08163v2), Table 3
48FLUX.2 [dev] · nf4 (bnb-4bit DiT + text encoder) Black Forest Labs49±2local~20GBhow to run › Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward ModelingLMArena leaderboard datasetGEditBench v2: A Human-Aligned Benchmark for General Image Editing
49P-Image-Edit Pruna48±3*api $0.010/imagereplicate.com ↗LMArena leaderboard dataset
51FLUX.2 [klein] 4B · bf16 + Qwen3 4B bf16 (sd.cpp) Black Forest Labs47±3*local~20GBhow to run › LMArena leaderboard datasetGEditBench v2: A Human-Aligned Benchmark for General Image Editing
52GLM-Image Z.ai47±6*api $0.015/image +local~39GBdocs.z.ai ↗GEditBench v2: A Human-Aligned Benchmark for General Image Editing
54Step1X-Edit v1.2 · bf16 StepFun46±2local (slow)~51GBhow to run › Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward Modelingstepfun-ai/Step1X-Edit-v1p2 model cardGEditBench v2: A Human-Aligned Benchmark for General Image EditingCREval: An Automated Interpretable Evaluation for Creative Image Manipulation under Complex Instructions (arXiv:2603.26174v1)
55Ideogram 3.0 Ideogram46±5*api $0.060/imagefal.ai ↗ServImage: An Image Generation and Editing Benchmark from Real-world Commercial Imaging Services (arXiv:2604.24023v2)
56Qwen-Image · gguf-q4_k_s Alibaba46±5*local~25GBhow to run › ServImage: An Image Generation and Editing Benchmark from Real-world Commercial Imaging Services (arXiv:2604.24023v2)
57Wan 2.5 Image-to-Image Preview Alibaba45±3*api $0.030/imagewww.alibabacloud.com ↗LMArena leaderboard dataset
58FLUX.1 Kontext [max] Black Forest Labs45±3*api $0.080/imagedocs.bfl.ai ↗LMArena leaderboard dataset
61GPT Image 1 Mini OpenAI42±3*api $0.011/imagedevelopers.openai.com ↗LMArena leaderboard datasetServImage: An Image Generation and Editing Benchmark from Real-world Commercial Imaging Services (arXiv:2604.24023v2)
62Step1X-Edit v1.1 · bf16 StepFun39±4local (slow)~51GBhow to run › stepfun-ai/Step1X-Edit-v1p2 model cardUnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched Edits
64FLUX.1 Kontext [pro] Black Forest Labs39±2api $0.040/imagedocs.bfl.ai ↗Meituan LongCat — LongCat-Image technical reportLMArena leaderboard datasetServImage: An Image Generation and Editing Benchmark from Real-world Commercial Imaging Services (arXiv:2604.24023v2)CREval: An Automated Interpretable Evaluation for Creative Image Manipulation under Complex Instructions (arXiv:2603.26174v1)
65FLUX.1 Kontext [dev] · gguf-q8_0 (sd.cpp) Black Forest Labs36±2local~28GBhow to run › Meituan LongCat — LongCat-Image technical reportLMArena leaderboard datasetEdit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward ModelingUnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched EditsGEditBench v2: A Human-Aligned Benchmark for General Image EditingCREval: An Automated Interpretable Evaluation for Creative Image Manipulation under Complex Instructions (arXiv:2603.26174v1)+2
67ChronoEdit-14B · bf16 + model offload (official script) NVIDIA34±3*local (slow)~34GBhow to run › Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward Modeling
68FLUX1.1 [pro] Black Forest Labs34±7*api $0.040/imagedocs.bfl.ai ↗ServImage: An Image Generation and Editing Benchmark from Real-world Commercial Imaging Services (arXiv:2604.24023v2)
69OmniGen2 · bf16 (official repo) BAAI34±2local~17GBhow to run › Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward ModelingUnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched EditsGEditBench v2: A Human-Aligned Benchmark for General Image EditingCREval: An Automated Interpretable Evaluation for Creative Image Manipulation under Complex Instructions (arXiv:2603.26174v1)BAAI — OmniGen2 technical reportMultiTextEdit: Benchmarking Cross-Lingual Degradation in Text-in-Image Editing (arXiv:2605.08163v2), Table 3+1
70BAGEL · nf4 (app.py --mode 2, bitsandbytes) ByteDance Seed33±2local~12GBhow to run › LMArena leaderboard datasetEdit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward ModelingUnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched EditsGEditBench v2: A Human-Aligned Benchmark for General Image EditingCREval: An Automated Interpretable Evaluation for Creative Image Manipulation under Complex Instructions (arXiv:2603.26174v1)BAAI — OmniGen2 technical report+3
71Stable Diffusion 3.5 Large · bf16 Stability AI33±7*local~20GBhow to run › ServImage: An Image Generation and Editing Benchmark from Real-world Commercial Imaging Services (arXiv:2604.24023v2)
72UNO · fp8 FLUX.1-dev base + UNO LoRA, CPU offload ByteDance32±7*local~16GBhow to run › BAAI — OmniGen2 technical report
74OmniGen (v1) · official OmniGen package, default settings BAAI30±7*†local~9GBhow to run › BAAI — OmniGen2 technical report
78Lumina-DiMOO · official inference_t2i.py (checkpoint as stored) Alpha-VLLM (Shanghai AI Lab)19±5*local (slow)~39GBhow to run › Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward Modeling
80OneCAT-3B · bf16 (official generate_txt2img.py) Meituan (OneCAT team)14±5*local~25GBhow to run › Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward Modeling
81InstructPix2Pix · fp16 UC Berkeley (Brooks, Holynski, Efros)10±4local~2.6GBhow to run › UnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched EditsEdit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward ModelingImagenWorld: Stress-Testing Image Generation Models with Explainable Human Evaluation on Open-ended Real-World TasksMultiTextEdit: Benchmarking Cross-Lingual Degradation in Text-in-Image Editing (arXiv:2605.08163v2), Table 3

+ 19 more don't fit your machine or budget · show them

81 of 81 catalogue models listed here · sources and licences

* new or thinly tested: ranked on too few benchmark results · † only the maker’s own numbers so far · ~RAM is estimated; full context may need more memory