comfy aimdo 0.2.2

[rattus128]

Comfy Aimdo 0.2.2 moves the cuda allocator hook from the cudart API to the cuda driver API on windows. This is needed to handle Windows+cu13 where cudart is statically linked.

Before:

Checkpoint files will always be loaded safely.
Total VRAM 12282 MB, total RAM 32607 MB
pytorch version: 2.10.0+cu130
Set vram state to: NORMAL_VRAM
Device: cuda:0 NVIDIA GeForce RTX 4070 : cudaMallocAsync
Using async weight offloading with 2 streams
Enabled pinned memory 14672.0
working around nvidia conv3d memory bug.
Using pytorch attention
aimdo: src-win/cuda-detour.c:35:ALL:aimdo_setup_hooks: cudart64_12.dll not foundNo working comfy-aimdo install detected. DynamicVRAM support disabled. Falling back to legacy ModelPatcher. VRAM estimates may be unreliable especially on Windows
Python version: 3.13.11 (tags/v3.13.11:6278944, Dec  5 2025, 16:26:58) [MSC v.1944 64 bit (AMD64)]
ComfyUI version: 0.15.0
ComfyUI frontend version: 1.39.16

After:

Total VRAM 8150 MB, total RAM 32691 MB
pytorch version: 2.10.0+cu130
Set vram state to: NORMAL_VRAM
Device: cuda:0 NVIDIA GeForce RTX 5060 : native
Using async weight offloading with 2 streams
Enabled pinned memory 14710.0
working around nvidia conv3d memory bug.
Using pytorch attention
aimdo: src-win/cuda-detour.c:77:INFO:aimdo_setup_hooks: found driver at 00007FFBAD8C0000, installing 4 hooksaimdo: src-win/cuda-detour.c:61:DEBUG:install_hook_entrys: hooks successfully installed
aimdo: src/control.c:65:INFO:comfy-aimdo inited for GPU: NVIDIA GeForce RTX 5060 (VRAM: 8150 MB)
DynamicVRAM support detected and enabled
Python version: 3.13.11 (tags/v3.13.11:6278944, Dec  5 2025, 16:26:58) [MSC v.1944 64 bit (AMD64)]
ComfyUI version: 0.14.1
ComfyUI frontend version: 1.39.16
[Prompt Server] web root: C:\Users\rattus\ComfyUI_windows_portable_nvidia\ComfyUI_windows_portable\python_embeded\Lib\site-packages\comfyui_frontend_package\static
Context impl SQLiteImpl.
Will assume non-transactional DDL.
Assets scan(roots=['models']) completed in 0.039s (created=0, skipped_existing=123, orphans_pruned=0, total_seen=123)
Starting server

Regression Tests:

For each of pyt 2.8+cu128 and pyt 2.12dev+cu130:
RTX5090, Linux, WAN 2.2 (multiple large models) ✅
RTX5090, Linux, Flux2 (linux offload) ✅
RTX5090, Linux, CPU@2GHZ Ace Step (CPU throttle) ✅

For each for pyt 2.8+cu128 and pyt2.10.0+cu130
Windows, RTX5060, WAN 2.1 640x400 ✅

Performance is roughly on par with non dynamic in the Windows WAN flow. A little faster (5-10s) on first runs due to load overlapping.

[coderabbitai]

📝 Walkthrough

Walkthrough

This pull request updates the comfy-aimdo dependency in requirements.txt from version 0.2.1 to 0.2.2. The change involves a single line modification in the requirements file, with one line added and one line removed. No other code modifications or functional changes are included in this update.

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title is concise and directly references the main change: upgrading comfy-aimdo to version 0.2.2. It clearly identifies the primary dependency update in the pull request.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Description check	✅ Passed	The PR description clearly explains the purpose of the version bump: moving CUDA allocator hooks from cudart API to driver API for Windows+cu13 compatibility.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[comfy-pr-bot]

Test Evidence Check

⚠️ Warning: Visual Documentation Missing

If this PR changes user-facing behavior, visual proof (screen recording or screenshot) is required. PRs without applicable visual documentation may not be reviewed until provided.

You can add it by:

• GitHub: Drag & drop media directly into the PR description
• YouTube: Include a link to a short demo