# welltop-cn/ComfyUI-TeaCache

Repository: https://github.com/welltop-cn/ComfyUI-TeaCache
Canonical: https://ross.abutalabs.com/products/comfyui-teacache
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2025-07-12T04:05:29+00:00

## Health v2 (maintenance only)
Score: 35/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 31, release rhythm 35, longevity 43
- inputs: {"age_days": 603, "days_push": 417, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1092, forks 66 (observed 2026-08-28T04:03:33.228823+00:00)

## What it is
A ComfyUI plugin integrating TeaCache, a training-free caching method that accelerates diffusion model inference by exploiting output differences across timesteps. It supports image, video, and audio diffusion models including FLUX, HiDream, Wan2.1, and HunyuanVideo with 1.5-2x speedups.

## Use cases
- speed up stable diffusion image generation in comfyui
- accelerate flux kontext inference
- faster wan2.1 video generation
- reduce diffusion model inference time without retraining
- speed up hunyuanvideo video synthesis
- make comfyui workflows run faster with caching

## When to choose
- you use ComfyUI and want faster diffusion generation with minimal quality loss
- you need training-free speedups for supported models like FLUX, HiDream, Wan2.1, or HunyuanVideo
- you want to trade slight quality degradation for up to 2x speed

## When to avoid
- you don't use ComfyUI for diffusion workflows
- you need maximum output quality with zero approximation
- your specific model version is not among the supported list

## Facets
- artifact type: plugin
- maturity: active
- function: machine-learning, llm-inference, image-processing, video-processing, caching
- domain: artificial-intelligence, image-processing, deep-learning
- platform: python, cross-platform
- tags: comfyui, diffusion-models, inference-acceleration, training-free, custom-nodes, stable-diffusion, flux, wan2-1, hunyuanvideo, video, gpu

## Member repositories
- welltop-cn/ComfyUI-TeaCache (main) score 35

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:33.228823+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:49:04.505277+00:00, confidence not recorded.
  - readme: https://github.com/welltop-cn/ComfyUI-TeaCache (fetched 2026-08-28T04:03:33.228823+00:00, sha 0f20ff7e8430)
- Data as of 2026-08-30T08:39:29.467469+00:00.