[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-75832":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":15,"subscribersCount":15,"size":15,"stars1d":14,"stars7d":16,"stars30d":17,"stars90d":15,"forks30d":15,"starsTrendScore":18,"compositeScore":19,"rankGlobal":9,"rankLanguage":9,"license":9,"archived":20,"fork":20,"defaultBranch":21,"hasWiki":22,"hasPages":20,"topics":23,"createdAt":9,"pushedAt":9,"updatedAt":24,"readmeContent":25,"aiSummary":26,"trendingCount":15,"starSnapshotCount":15,"syncStatus":27,"lastSyncTime":28,"discoverSource":29},75832,"Infinite-Canvas","hero8152\u002FInfinite-Canvas","hero8152","Supports comfyui\u002FAPI calls\u002Fmodelscope calls",null,"Python",1152,225,11,48,0,160,944,168,102.06,false,"main",true,[],"2026-06-12 04:01:19","# Infinite-Canvas\nSupports comfyui\u002FAPI calls\u002Fmodelscope calls\n\n详细教程：[https:\u002F\u002Fyoutu.be\u002F1y9ShTvgC_w](https:\u002F\u002Fyoutu.be\u002Fr_y_9ALr7fg)\n\n由于最近很多API网址关停，我找到一个稳定的网址：\n\nhttps:\u002F\u002Fapimart.ai\u002Fregister?aff=1uyAbb\n\n----\n\n【新增了version文件，我每次更新都会更新version的版本号，如果你下载version文件，打开项目后，导航栏的GitHub按键就会提示新版本，如果不想查看更新提示，就删除version文件】\n\n【A version file has been added. I update the version number with each update. If you download the version file, the GitHub button in the navigation bar will indicate the new version after opening the project. If you don't want to see update notifications, delete the version file.】\n\n----\n5\u002F13日更新：\n1. 修复了安装依赖的报错\n2. 增加了更便捷的API设置方式，现在可以全程在网页中设置，并且可以拉取模型一键添加\n3. LLM节点支持图片输入反推，可以使用modelscope的VL模型测试\n4. 增加了中英文切换\n5. 增加了自定义comfyui工作流的功能，可以自己设置需要的输入和要调整的参数，可以在无限画布的comfyui节点中调用。\n6. 增加了视频生成功能\n7. 修复了2k\u002F4k生成报错问题\n8. 新增了生成节点可以通过output前后连接一键生成\n\n5\u002F14日更新：\n1. 修复mac的一些bug\n2. modelscope支持lora调用\n3. 支持OpenAI协议和异步协议（生成失败不扣费）\n\n5\u002F15日更新：\n1. 增加了循环组件和计数功能，可以将节点循环\u002F并发运行N次，同时有提示词计数功能，可以设置提示词为：运行第二张卖点图。\n用法可以是：使用Gemini生成产品10个卖点提示词。用循环节点，增加提示词：运行第X张卖点图，输入给API节点，调用GPT生成，就可以一次性并发生成10个卖点图。\n2. 增加协议的验证按键，可以方便的验证自己的API平台是什么协议\n3. 修复了LLM节点的一些bug\n4. 上传了精简版的python，运行“安装依赖.bat”，完成之后，运行\"run.bat\"\n\n5\u002F18日更新：\n1. 修复了视频生成bug\n2. 增加了画笔编辑\n3. 修复了ccomfyui的并发bug\n   \n-----\n\nDetailed tutorial: [https:\u002F\u002Fyoutu.be\u002F1y9ShTvgC_w](https:\u002F\u002Fyoutu.be\u002Fr_y_9ALr7fg)\n\nDue to the recent shutdown of many API websites, I found a stable one:\n\nhttps:\u002F\u002Fapimart.ai\u002Fregister?aff=1uyAbb\n\n[Discount code (daxiong): 10% off your first top-up, invoice available.]\n\n\nMay 13th Update:\n\n1. Fixed dependency installation errors.\n\n2. Added a more convenient API setup method; settings can now be configured entirely through the webpage, and models can be added with a single click.\n\n3. LLM nodes support image input for reverse engineering; VL models from ModelScope can be used for testing.\n\n4. Added Chinese\u002FEnglish switching functionality.\n\n5. Added the ability to customize ComfyUI workflows; users can set their own inputs and adjust parameters, and these workflows can be invoked within ComfyUI nodes on an infinite canvas.\n\n6. Added video generation functionality.\n\n7. Fixed 2K\u002F4K generation error issues.\n\n8. Added the ability to generate videos with a single click by connecting output nodes before and after.\n\nMay 14th Update:\n\n1. Fixed some bugs on Mac.\n\n2. ModelScope supports LoRa calls.\n\n3. Supports OpenAI protocol and asynchronous protocol (no charge for generation failures).\n\nMay 15th Update:\n\n1. Added a loop component and counting function, allowing nodes to run concurrently N times. It also includes a prompt word counting function, where the prompt word can be set to: \"Run the second selling point image.\"\n\nUsage: Use Gemini to generate 10 selling point prompt words for a product. Use the loop node, add the prompt word: \"Run the Xth selling point image,\" input it to the API node, call GPT to generate, and you can generate 10 selling point images concurrently at once.\n\n2. Added a protocol verification button for easy verification of your API platform's protocol.\n\n3. Fixed some bugs in the LLM node.\n\n4. Uploaded a simplified version of Python. Run \"安装依赖.bat\", and then run \"run.bat\".\n   \n\nMay 18th Update:\n\n1. Fixed a video generation bug.\n\n2. Added brush editing.\n\n3. Fixed a concurrency bug in ccomfyui.\n\n\n\n\u003Cimg width=\"1696\" height=\"1350\" alt=\"b68e144c5b04a322bfd035da4d89aba3\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F0a6090fb-a8dd-4c3d-adee-b1f9233a2d91\" \u002F>\n\n   \n\u003Cimg width=\"1525\" height=\"1473\" alt=\"image\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F6f61fcf9-746c-425b-9e36-cfc8d252da7c\" \u002F>\n\n   \u003Cimg width=\"1261\" height=\"864\" alt=\"image\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F57f3e230-3134-488f-8179-d97e7d15383a\" \u002F>\n\u003Cimg width=\"1530\" height=\"858\" alt=\"image\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F9990e42d-22d5-4a10-a1e1-ad35a634edd2\" \u002F>\n\n\u003Cimg width=\"1735\" height=\"1400\" alt=\"image\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002Fd8328ff8-bbe0-4f1c-9ffa-7b56e8a1a51d\" \u002F>\n\u003Cimg width=\"2258\" height=\"969\" alt=\"image\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F4a752d99-885d-4ba9-8b86-91b495786b5c\" \u002F>\n\n\n\u003Cimg width=\"1531\" height=\"1374\" alt=\"image\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F0af79e38-0955-4740-9e65-5c9bb057f58c\" \u002F>\n\n\u003Cimg width=\"2196\" height=\"1040\" alt=\"image\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F6d823668-cde2-4836-8332-1858efe5f520\" \u002F>\n\u003Cimg width=\"2214\" height=\"771\" alt=\"image\" src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F52e10958-753f-45ba-a50e-3bbec27be436\" \u002F>\n","Infinite-Canvas 是一个支持 comfyui\u002FAPI 调用和 modelscope 调用的项目。其核心功能包括便捷的 API 设置、模型一键添加、LLM 节点的图像反推、中英文切换以及自定义 comfyui 工作流等。此外，还提供了视频生成、循环组件及计数功能，增强了项目的灵活性与实用性。该项目特别适用于需要通过网页界面进行复杂 AI 模型调用和管理的场景，如生成多张图片或视频内容、处理大规模数据集等。开发语言为 Python，持续更新并修复了多个平台上的已知问题，确保了良好的跨平台兼容性。",2,"2026-06-11 03:53:27","CREATED_QUERY"]