AI

人工智能学习路线

ComfyUI

部署

使用

资源

聚合API

应用

视频

音频

大模型在线应用

Chat

图像

  • 即梦 AI

未分类

  • GPT4共享账号 https://www.opkfc.com
  • https://gemini.google.com/app
  • NextChat
  • https://googlegemini.co
    • https://chat.googlegemini.co
  • https://www.coze.com
    • https://www.coze.cn
  • https://github.com/babaohuang/GeminiProChat
  • https://aitestkitchen.withgoogle.com/zh/tools/image-fx
  • https://ked.chatyy.org
  • perplexity AI 浏览器
  • https://metaso.cn
  • https://askmanyai.cn

图像创作

  • Lexica
  • https://dreamlike.art

通用大模型

文本

  • ChatGPT-4 可以使用文生图 $20 优选
  • Claude 3 整理长文本、数学、复杂问题。 $20
  • Gemini 对语音、图片理解能力更好 $20
  • Poe $20

图像

  • SD
  • Midjourney
  • Dall-E 3
  • Canva

视频

  • SORA
  • SVD 本地部署，目前的优选
  • Runway
  • Pika
  • Haiper 免费

音乐

  • Suno
  • Stable Audio
  • https://songgenerator.io/zh-CN

语音

  • https://voicechanger.im
  • https://texttospeech.im/zh-CN
  • https://github.com/2noise/ChatTTS
    • https://colab.research.google.com/github/Kedreamix/ChatTTS/blob/main/ChatTTS_infer.ipynb

数字人

  • HeyGen
  • D-iD
  • SadTalk 本地部署

开放模型

  • https://civitai.com

SD部署

Deepseek 部署

GPU pod

  • https://www.runpod.io
  • https://vast.ai
  • https://rundiffusion.com
  • https://www.coreweave.com

国内

  • https://www.autodl.com

一些AI应用

  • AI主播：https://luna.docs.ie.cx
  • 小白兔 AI：https://github.com/Baiyuetribe/paper2gui
  • https://www.liblib.art/
  • https://videosora.app/zh-cn

评测文章

  • https://www.163.com/dy/article/J4B8JQK60511DSSR.html
  • https://www.yeschat.ai 10次/天 gpt4

视频

  • https://phenaki.video/
  • https://makeavideo.studio/

Generate A.I Imagery From Referencing Your Own Poses - Dream Studio By Stable Diffusion Dearmstudio 基本使用

Create Art From Your Face With AI For Free | Dreambooth Tutorial | Stable Diffusion Guide

https://colab.research.google.com/github/ShivamShrirao/diffusers/blob/main/examples/dreambooth/DreamBooth_Stable_Diffusion.ipynb

PROMPT

portrait of 1girl by Bill Henson,depressed,tears,streaming tears,ruins,shadows,dramatic lighting,summer
contemporary,dark,expressionism,dystopia,industrial,<lora:MengX girl_Mix_V40:0.8>,<lora:add_detail:1>,

(worst quality:2),(low quality:2),(normal quality:2),lowres,bad anatomy,bad hands,normal quality,difConsistency_negative_v2,

masterpiece, photorealistic
 <lora:add_detail:1.5>
Negative prompt: easynegative, badhandv4, ng_deepnegative_v1_75t
Steps: 60, CFG scale: 7, Sampler: Euler a, Seed: 2631211893, VAE: vae-ft-mse-840000-ema-pruned.safetensors, Size: 2432x3200, Model: majicmixRealistic_v7, Version: v1.7.0, TI hashes: [object Object], Model hash: 7c819b6d13, ControlNet 0: [object Object], Denoising strength: 0.1, Ultimate SD upscale padding: 32, Ultimate SD upscale upscaler: 4x-UltraSharp, Ultimate SD upscale mask_blur: 8, Ultimate SD upscale tile_width: 512, Ultimate SD upscale tile_height: 512

overwatch, character portrait, concept art, intricate details, highly detailed by greg rutkowski, michael whelan and gustave dore

1girl, cityscape, skyscraper, sunset, highly detailed, sharp focus, extreme detail description, (vivid colors:0.6), cinematic lighting, extremely high-resolution details, photographic, realism pushed to extreme, fine texture, incredibly lifelike,side light, bright color

图像生成

之前为大家介绍了初代 AI 图像生成器 Disco Difussion、Dall·E 2、Midjounery，国产 AI 图像生成器 Tiamat，以及 4 款由谷歌、微软等大厂推出的 AI 图像生成器 Imagen、Parti、Make-A-Scene、NUWA-Infinity。

DreamStudio，Stability AI

https://beta.dreamstudio.ai/dream

介绍：https://www.bilibili.com/video/BV1cY4y1T7Zi

Dreamlike.art 官网直达： https://dreamlike.art/create （使用梯子速度更快，需注册账号）

模型地址： https://huggingface.co/dreamlike-art/dreamlike-diffusion-1.0

使用指南： https://dreamlike.art/guides

PromptHero 官网： https://prompthero.com/ (使用梯子速度更快)

urania.ai： https://www.urania.ai/top-sd-artists (使用梯子速度更快)

待梳理

更新时间：2025-05-09 12:59:41