whatmodel.sh / video / Wan 2.2 TI2V 5B
Wan 2.2 TI2V 5B
↗ share
+ contribute
Wan 2.2 TI2V 5B is an open-weights video model from Alibaba, released July 2025. Its published size is 5B parameters. Licence: apache-2.0.
$ how it compares
The video models ranked either side of Wan 2.2 TI2V 5B, out of 139 with a score, then more from Alibaba.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 79 |  Cosmos-Predict2-2B-Video2World | 48±8 | comparison only | ||
| 80 | Wan 2.1 14B | 47±2 | api $0.24/sec +local | ||
| 81 | Wan 2.2 TI2V 5B | 47±2 | api price unknown +local | ||
| 82 |  Kandinsky 5.0 T2V Lite | 47±2 | local | ||
| 83 |  LTX-2 19B Distilled | 47±8 | api $0.020/sec +local | ||
| more from Alibaba | |||||
| 2 | Wan 3.0 Prime | 80±7 | api $0.14/sec | ||
| 4 | Wan 3.0 | 79±2 | api $0.10/sec | ||
| 9 | HappyHorse 1.1 | 76±3 | api $0.14/sec | ||
$ what it's good at
text to video▆▄▃▄▆▅█▃▂▂▂▄ #41 of 68 · 40% of its score · 4 tests
Trails Hailuo 01 (MiniMax video-01 / I2V-01), Cosmos-1.0-Diffusion-7B-Text2World, LTX-2 19B Distilled here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Cosmos HUE human evaluation, text-to-video (Cosmos 3 report) | 80.8% ✎ | █▄▄··▄▄▄▄▄·▄ beats 0/9 | 21% |
| VABench T2AV micro Visual QA | 63.6% ✎ | ▄····▄····▄█ beats 0/4 | 12% |
| Sci-VBench testmini (expert human rating) | 44.4% ✎ | ▅·▃█▃·▅▃▅▃▃▃ beats 3/14 | 7.0% |
| VBench T2V total (MobileWan 480p ablation) | 81.3% ✎ | no other model yet | — |
image to video▁▂·▂▅█▅▄▅▆▃▄ #43 of 50 · 38% of its score · 4 tests
Beats Wan 2.2 A14B, LTX-2 here, though they rank higher overall. Trails Hailuo 01 (MiniMax video-01 / I2V-01) here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Cosmos HUE human evaluation, image-to-video (Cosmos 3 report) | 80.4% ✎ | ▄·█··▄··▄▄▄█ beats 0/8 | 16% |
| Target-Bench v2 (VGGT planning score, I2V) | 8.4% ✎ | ▃·▃···▃▃█·▃▃ beats 1/8 | 11% |
| PAI-Bench-G physical-AI video generation (Overall) | 81.3% ✎ | ▃······▅·█▅█ beats 7/10 | 6.6% |
| What-If World causal-intervention video generation (APEO paired average, Gemini 3.1 Pro judge) | 28.4% ✎ | █·▄····▄···█ beats 1/5 | 4.5% |
other tests3 not in a task · 20% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| VideoArgus-Bench TI2V (rubric-grounded score) | 78.2% ✎ | █··▄▄▄▄····▄ beats 2/6 | 8.8% |
| PhyGround physics human evaluation (Overall, TI2V) | 52.6% ✎ | █▄▄··▄···▄·█ beats 2/7 | 5.8% |
| MechVerse mechanical-motion I2V (WorldModelBench-judge sum) | 65.6% ✎ | ▂···▃······█ beats 1/5 | 5.5% |
$ where to run it
Through an API
api · price unknown · fal.ai ↗
Other ways to run it
version
- Run it as published · needs ~10.7 GB · from github.com/leejet
sd-cli -M vid_gen --diffusion-model models/Wan2.2-TI2V-5B-Q4_K_M.gguf --vae models/wan2.2_vae.safetensors --t5xxl models/umt5-xxl-encoder-Q4_K_M.gguf -p "a lovely cat" --cfg-scale 6.0 --sampling-method euler -v -n "色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走" -W 480 -H 832 --diffusion-fa --offload-to-cpu --video-frames 33 --flow-shift 3.0
Weights and documentation: huggingface.co/Wan-AI ↗
$ how the score was worked out · Cosmos 3: Omnimodal World Mod… 37%, VABench: A Comprehensive Benc… 12%, Target-Bench: Can Video World… 11% and 6 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
not counted (1)
| MobileWan: Closing the Quality Gap for Mobile Video Diffusion | VBench T2V total (MobileWan 480p ablation) [correct] | 81.3% | fewer than three models have this test yet, or this source ran it for this model alone |
$ published speed measurements · 1
Reference results for the hardware, engine and settings below. They aren't verified speeds of the downloads above or predictions for your machine.
| NVIDIA 4090 | bf16 (--convert_model_dtype; shared config) | 534.7 s/video | Wan reference implementation · TI2V-5B T2V; 720P; 1 GPU; --offload_model True; no prompt extension; --convert_model_dtype; --t5_cpu; mean after warmup PNG time column manually visually verified. Combined GPU labels preserved; no claim that the run was specifically the H100 rather than H800. CPU model offload is explicit in settings; not misclassified as no offload. | source ↗ |
$ 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