AI Pose Gallery 中文版
Chinese pose, outfit and portrait reference library for AI art
Description
Drawing characters with AI is rarely about style, it's about references: how to pose, what to wear, what face and hairstyle feel right, and the Japanese original sites mean a lot of browsing and translating prompts. This project merges three reference galleries and fully localizes them into Chinese: poses, outfits and portraits, 4,345 entries and nearly 9,000 local images you can combine into a character recipe.
Beyond browsing it includes a character studio that merges the three picks into one board and sends it straight to Gemini, OpenAI, Alibaba Bailian, Volcano Ark or a local ComfyUI, keeping history in the browser; the repo is also a ComfyUI custom node plugin. Use the hosted site or serve it statically yourself. The original galleries come from a Japanese author and the Chinese edition is maintained by Anhao.
Three reference libraries: poses by system, action category and angle; outfits in 8 seasonal and style groups; portraits by face reference or styling, hair length and 51 hairstyles.
Chinese search and lightbox: live keyword search with full size previews and downloads.
Character studio: pick one from each library plus any custom references, with 12 style presets and auto assembled Chinese prompts.
Multi provider generation: Google Gemini, OpenAI GPT Image, Alibaba Qwen, Volcano Doubao and local ComfyUI.
ComfyUI plugin: clone into custom_nodes to get pose, outfit, portrait and combined recipe nodes with a visual picker.
Local storage: history in browser IndexedDB, and generation continues when you switch pages.
Beyond browsing it includes a character studio that merges the three picks into one board and sends it straight to Gemini, OpenAI, Alibaba Bailian, Volcano Ark or a local ComfyUI, keeping history in the browser; the repo is also a ComfyUI custom node plugin. Use the hosted site or serve it statically yourself. The original galleries come from a Japanese author and the Chinese edition is maintained by Anhao.
Features
Three reference libraries: poses by system, action category and angle; outfits in 8 seasonal and style groups; portraits by face reference or styling, hair length and 51 hairstyles.
Chinese search and lightbox: live keyword search with full size previews and downloads.
Character studio: pick one from each library plus any custom references, with 12 style presets and auto assembled Chinese prompts.
Multi provider generation: Google Gemini, OpenAI GPT Image, Alibaba Qwen, Volcano Doubao and local ComfyUI.
ComfyUI plugin: clone into custom_nodes to get pose, outfit, portrait and combined recipe nodes with a visual picker.
Local storage: history in browser IndexedDB, and generation continues when you switch pages.
