Wan2.2-TI2V-5B
Open-weight 5B text-and-image-to-video model. Working set 34.2 GB: transformer 20.0, umT5 text encoder 11.4, VAE 2.8.
Runssource: Hugging FacePythonApache-2.0revision 921dbaf3f167
Python, Apache-2.0 licensed. The project labels itself: text to video, video generation and diffusion.
Wan2.2-TI2V-5B runs. An Argusic agent installed it in 63 minutes and hit 4 errors and fixed every one without help on a clean machine, and the whole session was recorded.
Generated a 1280x704 H.264 video, 41 frames, 1.71 seconds at 24 fps, 2.8 MB, decoded with zero errors.
What the agent ended up with on a clean machine, in its own words. How this is measured
Measured by Argusic on a fresh machine every time. Every number links to its evidence. Argusic Score 100 of 100 (the mean of this project's run scores).
At a glance
Subject data from Hugging Face, linked at the top of this page. Test data by Argusic (CC BY 4.0); every number links to a run page with the full log, the recording, and their sha256 hashes.
What was measured
From the run that decided this verdict. Every line is a recorded value, not an estimate.
What went wrong, in the agent's own words
The agent's own words, worst first, never rewritten into advice.
- flash-attn cannot compile without CUDA_HOME/nvcc15 minutes
- CUDA OOM at 121 frames (VAE decode)8 minutes
- Missing packages: einops, decord, librosa, peft, packaging3 minutes
- librosa 1.0.0 incompatible with numpy<22 minutes
- Install time
- 63 minutes
- Cold machine to finish
- 65 minutes
- Errors hit and fixed
- 4 hit, 4 fixed with no human help
- How the result was proved
- ffmpeg -v error -i out.mp4 -f null - exited 0, no errors. ffprobe reports coded_width=1280, coded_height=704, nb_frames=41, duration=1.708333s, 24 fps, H.264.
- Model tokens used
- 584,130
- Exact commit tested
- 921dbaf3f167
- Container image
- sha256:cdd920bce783
What was not measured
The other half, so you can see exactly how far this evidence goes. A gap is printed as a gap and never filled in.
- Peak GPU memory
- GPU memory sampling was added on 2026-09-07; this run predates it
- Behaviour over a long run
- each attempt uses the smallest documented settings and a fixed time budget, so this measures whether it runs, not how it holds up under load
The end of the recording
The last lines of the terminal session that decided this verdict, exactly as they were recorded.
- **Test depth**: **tier1_real** — real output produced and decoded successfully Two notable fixes were needed: (1) patching `wan/modules/attention.py` `flash_attention()` to fall … tokens used 584,130 The result line has been printed above. Let me summarize the test run for you: **Wan2.2-TI2V-5B test completed successfully** - **GPU**: RTX 4090 (24 GB VRAM), 120 GB free disk before download - **Install**: all dependencies and weights (23 files, ~34 GB total) installed in ~13 minutes; runt… - **Inference**: ran the model card's documented command for 24 GB GPUs (`--size 1280*704 --offload… - **Output**: 1280×704 H.264 video, 41 frames, 1.71s at 24 fps, 2.8 MB — decoded with zero errors v… - **Test depth**: **tier1_real** — real output produced and decoded successfully Two notable fixes were needed: (1) patching `wan/modules/attention.py` `flash_attention()` to fall …
Replay the whole session, every command from a clean machine to this point.
How it was tested
One agent, one container, no human help. It installs the project the way its own documentation says to, runs it, and fixes what breaks. Everything below is recorded as it happened: the terminal session, the log and the exact commit. The full method.
Strengths and limits
Measured facts, not opinions. How this is written.
What went well
- Reached a running state on a clean machine, with the session recorded.
- Installed in 63 minutes.
- Recovered from all 4 errors without a human stepping in, which says the failures are documented well enough to solve.
- Licensed Apache-2.0, as reported by its host.
- Measured 2 times, so the result is not a one-off.
Run history
Topics (from Hugging Face)
Embed the badge
Markdown for the project README. It links back here; terms on the terms page.
[](https://argusic.com/subject/wan2-2-ti2v-5b)Questions
- Does Wan2.2-TI2V-5B run?
- Yes. Wan2.2-TI2V-5B runs. Argusic installed and launched it on a clean machine in 63 minutes, hitting 4 errors on the way, and recorded the session.
- How did Argusic test Wan2.2-TI2V-5B?
- On a fresh, disposable machine, with every command recorded and the repository pinned at commit 921dbaf3f167. 2 attempts are recorded, and the full method is on the methodology page.
- What did it cost to test Wan2.2-TI2V-5B?
- The run that produced this verdict cost $0.93: real compute and model cost, not a list price per million tokens. It is recorded on that run's page.
- How long does Wan2.2-TI2V-5B take to install?
- 63 minutes in the run that decided this verdict, from a clean machine to a running state. That is one measured attempt, not an average.
- Does Wan2.2-TI2V-5B need a GPU?
- Argusic ran it on a GPU machine, so this page cannot say whether it starts without one.
- What goes wrong when installing Wan2.2-TI2V-5B?
- 4 things broke in the recorded run, and 4 were fixed without human help. Each one, and the time it cost, is listed on this page.
- Where is the evidence for Wan2.2-TI2V-5B?
- All 2 recorded runs are on this page, each linking to its full log and terminal recording, stored with a sha256 fingerprint so it cannot be quietly altered.