Grok Imagine - 2
Create sharp, photorealistic AI images from text prompts with Grok Imagine 2. Try the independent Grok Imagine Image V2.0 generator with 32 signup credits.
UpscaleAI
Upscale and enhance any photo or video with AI. Restore old photos; sharpen anime, faces and text; take images to 8K and video to 4K with Topaz models.
Free Image Generator AI
Free AI image generator that turns text prompts into stunning AI art in seconds. Create images from text — no signup, no credit card, no watermark.
pxdmd
online AI image editor. Remove image backgrounds, erase watermarks and unwanted objects, change backgrounds, and tint photos — all in seconds with 3 free credits. No design skills needed.
ApiFlux
ApiFlux is an AI router with native Anthropic, OpenAI & Gemini APIs — one key for 100+ models, automatic failover, transparent per-token pricing.
ai influencer generator
Create your AI influencer, save their identity, and make consistent photos, videos, and UGC-style content. Start with 30 Credits and see costs before generating.
PictureMaker
Create AI art, logos, avatars, and more with PictureMaker's AI Picture Maker. Try a profile picture maker or cartoon picture maker free today.
Grok Image 2.0
Create and edit images online with Grok Image 2.0. Generate visuals from text or refine a reference image for products, ads, posters, social content, and more.
Graded Prompts
Buy AI prompts a person ran on the real model before sale. Image, text and video prompts for Midjourney, ChatGPT, Kling and 20+ models. Free prompts included.
image mcp server
Image MCP Server gives Claude Code, Cursor and Copilot AI image tools for websites: hero images, OG cards, product cutouts, favicons. MCP or REST API.
Get3W
Get3W 是面向开发者和企业的 AI 模型聚合平台，通过统一 API 一站接入来自 Google、OpenAI、ByteDance、Alibaba、Kling、Midjourney、Runway、MiniMax、Black Forest Labs 等 20 多家厂商的 700+ 模型，能力覆盖视频、图像、音频、3D、全景与对话生成。
PixelHacker
PixelHacker 是华中科技大学和 VIVO AI Lab联合推出的图像修复（Image Inpainting）模型。基于引入潜在类别引导（Latent Categories Guidance, LCG）范式，分别对前景和背景进行编码，基于线性注意力将特征注入去噪过程中，实现结构和语义的一致性。
D
D-DiT（Dual Diffusion Transformer）是卡内基梅隆大学、耶鲁大学和字节跳动Seed实验室推出的多模态扩散模型，能统一图像生成和理解任务。模型结合连续图像扩散（流匹配）和离散文本扩散（掩码扩散），基于双向注意力机制同时训练图像和文本模态。
Ev
Ev-DeblurVSR是中国科学技术大学、合肥综合性国家科学中心人工智能研究所和新加坡国立大学联合推出的视频画面增强模型，能从低分辨率且模糊的视频输入中恢复出高分辨率、清晰的视频。
Step1X
Step1X-Edit 是阶跃星辰团队推出的通用图像编辑框架，能缩小开源图像编辑模型与闭源模型（如 GPT-4o 和 Gemini2 Flash）之间的性能差距。Step1X-Edit结合多模态大语言模型（MLLM）和扩散模型，基于处理参考图像和用户的编辑指令，提取潜在嵌入并生成目标图像。
GigaTok
GigaTok 是用于自回归图像生成的视觉分词器，参数量达 30 亿。通过语义正则化技术，将分词器特征与预训练视觉编码器（如 DINOv2）的语义特征对齐，有效约束潜在空间复杂度，解决了视觉分词器在扩展时重建质量与生成质量之间的矛盾。
InfiniteYou
InfiniteYou（InfU）是字节跳动智能创作团队推出的基于扩散变换器（Diffusion Transformers，如 FLUX）的身份保持图像生成框架。基于 InfuseNet 将身份特征注入扩散模型，增强身份相似度，保持图像生成能力。
MagicArena
MagicArena 是字节跳动推出的国内首个视觉生成大模型对战平台。用户输入文字，能调用不同模型（如 如 Midjourney、FLUX、可灵、海螺、即梦等）生成图片或视频，进行比较、投票。
ChatLLM Teams
An AI assistant platform providing access to state-of-the-art Large Language Models (LLMs), web search, and image generation for individuals and teams.
Live3D AI Face Swap
Online AI face swap for photos and videos with no-login workflow, watermark-free results, and daily free video swaps.