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

Best image-to-video models for a 32GB Mac

A leaderboard of the best image-to-video models for a Mac with 32GB of memory. The best one that runs on it right now is Wan 2.2 TI2V 5B. Models are ranked on human votes and tests that turn an image into a video. Models that run on it show the download that fits and the command to run it.

46 models · scores combine 16 benchmarks · updated how it works · find a leaderboard · benchmark your machine
task[all tasks][text to video][image to video][video editing]

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

Models for 32GB RAM · Apple Silicon → ~21GB fast · reset
+ more
chip[M1 Pro][M1 Max][M2 Pro][M2 Max][M4][M5][M6 (24/32GB)]

Detect your exact chip or graphics card:

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

curl -G whatmodel.sh/video -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/video -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/video?ram=32&gpu=apple"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]
price[local][$0.02][$0.05][$0.1][$0.25][any]max per sec

detecting your OS…

Score vs API price

#modeltask score↓where↓speed↓RAM↓get itsources
#modeltask score↓where↓speed↓RAM↓get itsources
1Gemini Omni 1.1 Flash Google82±2api $0.10/secai.google.dev ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard datasetRapidata/camera-movement published dataset-card rankings
2MiniMax H3 MiniMax80±2api $0.13/sec~52GBfal.ai ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard datasetVGI-Bench: Probing Visual Intelligence in Video Generation ModelsRapidata/camera-movement published dataset-card rankings
3Gemini Omni Flash Google77±2*api $0.10/secai.google.dev ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard dataset
4Dreamina Seedance 2.0 ByteDance Seed77±2api $0.30/secfal.ai ↗Rapidata/world-model-physics published dataset-card rankingsWorldReasonBench: Human-Aligned Stress Testing of Video Generators as Future World-State Predictors (arXiv v1)LMArena leaderboard datasetVGI-Bench: Probing Visual Intelligence in Video Generation ModelsFrom Evaluation to Enhancement: Benchmarking and Improving Think-with-Video Reasoning for Video Generative Models (VWG-Bench)What-If World: A Causal Benchmark for General World Models in Embodied Scenarios (arXiv:2605.27589v1), Table 2
5Dreamina Seedance 2.5 ByteDance75±2*api $0.30/secdocs.dev.runwayml.com ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard dataset
6FLUX 3 Video Black Forest Labs75±2api $0.17/secdocs.bfl.ai ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard datasetRapidata/camera-movement published dataset-card rankings
7Wan 3.0 Alibaba73±2api $0.10/secwww.alibabacloud.com ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard datasetRapidata/camera-movement published dataset-card rankings
9Wan 2.2 I2V Flash Alibaba70±5*api $0.036/secwww.alibabacloud.com ↗Target-Bench: Can Video World Models Achieve Mapless Path Planning with Semantic Targets?
10Grok Imagine Video 1.5 xAI70±2api $0.080/secdocs.x.ai ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard datasetRapidata/camera-movement published dataset-card rankings
11Kling 3.0 Kuaishou (Kling AI)69±7*api $0.084/seckling.ai ↗VGI-Bench: Probing Visual Intelligence in Video Generation Models
12Grok Imagine Video xAI67±2api $0.050/secdocs.x.ai ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard datasetWhat-If World: A Causal Benchmark for General World Models in Embodied Scenarios (arXiv:2605.27589v1), Table 2Physics-IQ Verified
13HappyHorse 1.0 Alibaba-ATH67±2*api $0.14/secreplicate.com ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard dataset
14Wan 2.2 I2V Plus Alibaba66±5*api $0.10/secwww.alibabacloud.com ↗Target-Bench: Can Video World Models Achieve Mapless Path Planning with Semantic Targets?
15Kling 3.0 Pro Kuaishou66±2api $0.11/secfal.ai ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard datasetRapidata/camera-movement published dataset-card rankings
16Veo 3.1 Fast Google64±2api $0.10/secai.google.dev ↗Rapidata/world-model-physics published dataset-card rankingsWorldReasonBench: Human-Aligned Stress Testing of Video Generators as Future World-State Predictors (arXiv v1)LMArena leaderboard datasetTarget-Bench: Can Video World Models Achieve Mapless Path Planning with Semantic Targets?
17Veo 3.1 Google64±2api $0.40/secai.google.dev ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard datasetTarget-Bench: Can Video World Models Achieve Mapless Path Planning with Semantic Targets?VGI-Bench: Probing Visual Intelligence in Video Generation ModelsCosmos 3: Omnimodal World Models for Physical AI (arXiv:2606.02800v4)From Evaluation to Enhancement: Benchmarking and Improving Think-with-Video Reasoning for Video Generative Models (VWG-Bench)+3
18Wan 2.7 Alibaba63±2api $0.10/secwww.alibabacloud.com ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard datasetVGI-Bench: Probing Visual Intelligence in Video Generation Models
19Runway Gen-4.5 Runway62±7*api $0.12/secdocs.dev.runwayml.com ↗VGI-Bench: Probing Visual Intelligence in Video Generation Models
20Wan 2.6 Alibaba61±2api $0.10/secwww.alibabacloud.com ↗Rapidata/world-model-physics published dataset-card rankingsWorldReasonBench: Human-Aligned Stress Testing of Video Generators as Future World-State Predictors (arXiv v1)LMArena leaderboard dataset
21Kling 2.6 Pro Kuaishou (Kling AI)60±2*api $0.070/secfal.ai ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard dataset
22Vidu Q3 Pro ShengShu (Vidu)60±2*api $0.10/secplatform.vidu.com ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard dataset
24HappyHorse 1.1 Alibaba57±15*api $0.14/secwww.alibabacloud.com ↗Rapidata/camera-movement published dataset-card rankings
26PixVerse V5.6 PixVerse56±2*api $0.045/secdocs.platform.pixverse.ai ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard dataset
27Wan 2.5 Preview Alibaba56±2api $0.10/secwww.alibabacloud.com ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard datasetTarget-Bench: Can Video World Models Achieve Mapless Path Planning with Semantic Targets?From Evaluation to Enhancement: Benchmarking and Improving Think-with-Video Reasoning for Video Generative Models (VWG-Bench)
28Wan 2.1 I2V Plus Alibaba53±6*api price unknownwww.alibabacloud.com ↗Target-Bench: Can Video World Models Achieve Mapless Path Planning with Semantic Targets?
32Seedance 1.0 Pro ByteDance Seed50±4*api $0.052/secdocs.byteplus.com ↗LMArena leaderboard dataset
33Kling 2.5 Turbo Pro Kuaishou49±4api $0.070/secfal.ai ↗LMArena leaderboard datasetFrom Evaluation to Enhancement: Benchmarking and Improving Think-with-Video Reasoning for Video Generative Models (VWG-Bench)V-ReasonBench: Toward Unified Reasoning Benchmark Suite for Video Generation Models
34Wan 2.6 Flash Alibaba49±8*api $0.025/secwww.alibabacloud.com ↗From Evaluation to Enhancement: Benchmarking and Improving Think-with-Video Reasoning for Video Generative Models (VWG-Bench)
35Hailuo 2.3 MiniMax48±4*api $0.047/secplatform.minimax.io ↗LMArena leaderboard dataset
37Wan 2.2 I2V A14B Alibaba48±2api $0.080/sec~31GBfal.ai ↗WorldReasonBench: Human-Aligned Stress Testing of Video Generators as Future World-State Predictors (arXiv v1)PAI-Bench: A Comprehensive Benchmark For Physical AI (arXiv:2512.01989v1)CRONOS: Benchmarking Counterfactual Physical Consistency in Video Models (arXiv:2605.23699v1), Table 1Cosmos 3: Omnimodal World Models for Physical AI (arXiv:2606.02800v4)VGI-Bench: Probing Visual Intelligence in Video Generation ModelsPhysics-IQ Verified
38Vidu Q2 Turbo ShengShu (Vidu)46±4*api $0.050/secfal.ai ↗LMArena leaderboard dataset
39P-Video Pruna46±4*api $0.020/secreplicate.com ↗LMArena leaderboard datasetPhysics-IQ Verified
40Hailuo 02 MiniMax45±4*api $0.047/secplatform.minimax.io ↗LMArena leaderboard datasetV-ReasonBench: Toward Unified Reasoning Benchmark Suite for Video Generation Models
42Kling 2.1 Master Kuaishou (Kling AI)45±4*api $0.28/secreplicate.com ↗LMArena leaderboard dataset
43Seedance 1.5 Pro ByteDance Seed45±2api $0.052/secfal.ai ↗Rapidata/world-model-physics published dataset-card rankingsLMArena leaderboard datasetCosmos 3: Omnimodal World Models for Physical AI (arXiv:2606.02800v4)What-If World: A Causal Benchmark for General World Models in Embodied Scenarios (arXiv:2605.27589v1), Table 2
44Kling 2.1 Standard Kuaishou (Kling AI)44±4*api $0.050/secreplicate.com ↗LMArena leaderboard dataset
47HunyuanVideo 1.5 (I2V) Tencent44±3api $0.075/secfal.ai ↗WorldReasonBench: Human-Aligned Stress Testing of Video Generators as Future World-State Predictors (arXiv v1)Cosmos 3: Omnimodal World Models for Physical AI (arXiv:2606.02800v4)VGI-Bench: Probing Visual Intelligence in Video Generation ModelsWhat-If World: A Causal Benchmark for General World Models in Embodied Scenarios (arXiv:2605.27589v1), Table 2Physics-IQ Verified
49LongCat-Video Meituan43±3*api $0.040/secfal.ai ↗WorldReasonBench: Human-Aligned Stress Testing of Video Generators as Future World-State Predictors (arXiv v1)
55Hailuo 01 (MiniMax video-01 / I2V-01) MiniMax41±5*api $0.083/secfal.ai ↗SkyReels-V2: Infinite-length Film Generative Model
57HunyuanVideo 1.5 Tencent40±4*api $0.075/sec~34GBfal.ai ↗LMArena leaderboard dataset
58Wan 2.2 TI2V 5B · fp16 + umt5 q8_0 (sd.cpp) Alibaba40±4local~21GBhow to run › Cosmos 3: Omnimodal World Models for Physical AI (arXiv:2606.02800v4)Target-Bench: Can Video World Models Achieve Mapless Path Planning with Semantic Targets?PAI-Bench: A Comprehensive Benchmark For Physical AI (arXiv:2512.01989v1)What-If World: A Causal Benchmark for General World Models in Embodied Scenarios (arXiv:2605.27589v1), Table 2
61Wan 2.1 I2V Turbo Alibaba38±7*api price unknownwww.alibabacloud.com ↗Target-Bench: Can Video World Models Achieve Mapless Path Planning with Semantic Targets?
64Wan 2.2 A14B Alibaba37±4*api $0.080/sec~31GBfal.ai ↗LMArena leaderboard dataset
71LTX-2.3 Lightricks23±3*api $0.040/sec~34GBdocs.ltx.io ↗WorldReasonBench: Human-Aligned Stress Testing of Video Generators as Future World-State Predictors (arXiv v1)
72Runway Gen-4 Turbo Runway23±4*api $0.050/secdocs.dev.runwayml.com ↗LMArena leaderboard dataset
74Pika 2.2 Pika16±4*api $0.040/secfal.ai ↗LMArena leaderboard dataset

+ 28 more don't fit your machine or budget · show them

74 of 74 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