Daily Digest – 2026-05-22

cawa-93/vite-electron-builder

Building a desktop app? This boilerplate delivers a secure Electron setup with Vite and your favorite JS framework. (TypeScript, 3k★)

★ 2,979

apptension/saas-boilerplate

Stop rebuilding auth, payments, and admin; this SaaS boilerplate gets your product live on AWS faster. (TypeScript, 2.9k★)

★ 2,888

chibat/chrome-extension-typescript-starter

Building a Chrome Extension? This starter gets you going with TypeScript and React without fighting webpack. (TypeScript, 2.8k★)

★ 2,824

codercup/unibest

For cross-platform Chinese app development, this UniApp starter with Vue3 and Vite makes it actually enjoyable. (TypeScript, 2.8k★)

★ 2,816

spliit-app/spliit

Tired of Splitwise? This open-source alternative on Next.js lets you share expenses without the data privacy concerns. (TypeScript, 2.7k★)

★ 2,654

zhaochen0110/Awesome_Think_With_Images

Deep dive into how Large Vision Language Models reason with images; essential reading for multimodal AI builders. (Multi, 1.4k★)

★ 1,434

alan345/AI-Fullstack-SaaS-Boilerplate

Kickstart your AI SaaS with this fullstack boilerplate, complete with auth, payments, and agentic AI integrations. (TypeScript, 1.4k★)

★ 1,394

apocas/restai

Host your own AI-as-a-Service platform, supporting local LLMs, embeddings, and even image/audio generation. (Python, 494★)

★ 494

metorial/metorial-platform

Build and scale custom AI agents easily; this platform is designed for multi-agent workflows. (TypeScript, 206★)

★ 206

digitalocean/sea-notes-saas-starter-kit

Launch a notes SaaS fast with this Next.js kit, including AI features, payments, email, and storage. (TypeScript, 138★)

★ 138

tremorlabs/template-dashboard-oss

Need a quick dashboard? This Next.js template by Tremor gives you a solid base for visualizing data. (TypeScript, 504★)

★ 504

argus-sight/refund-swatter-lite

If you're selling on Apple App Store, this tool helps prevent fraudulent refunds by reporting consumption data. (TypeScript, 169★)

★ 169

BeyteFlow/ReadmeGenAI

Stop dreading READMEs; this AI-powered generator creates professional ones from your GitHub repo. (TypeScript, 83★)

★ 83

KlingAIResearch/LivePortrait

Animate static portraits into dynamic videos; perfect for engaging content or historical recreations. (Python, 18.2k★)

★ 18,235

camenduru/stable-diffusion-webui-colab

Run Stable Diffusion WebUI on Colab effortlessly, making AI art generation accessible without local setup. (Jupyter Notebook, 15.9k★)

★ 15,943

Wan-Video/Wan2.2

Cutting-edge open-source video generation models for creating high-quality, large-scale video content. (Python, 15.5k★)

★ 15,535

duixcom/Duix-Avatar

Clone digital humans and generate AI avatars offline; a powerful toolkit for personalized video creation. (C, 12.9k★)

★ 12,883

google/gemini-3-pro

Access Google's Gemini 3 Pro for advanced reasoning and multi-modal tasks via Replicate's API. (1.2M runs)

google/gemini-3.1-pro

The latest Gemini 3.1 Pro via Replicate, offering improved reasoning and better thinking capabilities. (586K runs)

sync/lipsync-2-pro

Get studio-grade lipsync for your videos quickly and easily through an API call. (35K runs)

xai/grok-imagine-image-quality

Grok's higher-quality image model via API, delivering sharper details and better text rendering. (19K runs)

moonshotai/kimi-k2-thinking

Moonshot AI's Kimi K2 Thinking model for complex reasoning and conversational AI applications. (5K runs)

yi-coder

Need a local code assistant? Yi-Coder offers strong coding performance without huge hardware demands. (953.3K pulls)

athene-v2

A 72B model for complex code completion, math, and log analysis; runs locally via Ollama. (553K pulls)

nous-hermes2-mixtral

A powerful Mixtral-based model for complex reasoning and creative tasks, accessible locally via Ollama. (552.8K pulls)

codeup

Get solid code generation locally with this Llama2-based model via Ollama. (552.7K pulls)

qwen3-next

The new Qwen3-Next model offers strong performance and efficiency for local reasoning and tools via Ollama. (531.7K pulls)

proxy-agent

Need to route HTTP/S traffic through various proxies? This package simplifies agent configuration. (28.0M/wk)

pac-proxy-agent

If your network uses PAC files for proxy configuration, this makes `http.Agent` work seamlessly. (26.8M/wk)

superagent

A trusty HTTP client for Node.js and browsers, offering a clean, fluent API for requests. (20.3M/wk)

puppeteer-core

Control headless Chrome or Edge programmatically; essential for web scraping and browser automation. (18.0M/wk)

puppeteer

Automate Chrome actions, scrape websites, or generate PDFs programmatically with this robust API. (10.4M/wk)

openai

The official Python library for interacting with OpenAI's API, making LLM integration straightforward. (74.6M recent downloads)

sentence-transformers

Easily generate text embeddings, build semantic search, or improve RAG systems with this library. (6.4M recent downloads)

dalle-mini/dalle-mini

Still good for a quick laugh, DALL-E Mini generates wild images from your text prompts. (static, 5685 likes)

AP123/IllusionDiffusion

Play with AI-generated optical illusions; a fun way to create trippy images from text. (gradio, 5403 likes)

Wan-AI/Wan2.2-Animate

Animate images into videos; a hands-on demo of the Wan2.2 video generation model. (gradio, 5120 likes)

facebook/MusicGen

Generate music from text prompts; useful for background tracks or quick sound ideas. (gradio, 5067 likes)

black-forest-labs/FLUX.1-schnell

A fast inference demo of the FLUX.1 model, quickly generating images from text prompts. (gradio, 5064 likes)

Qwen/Qwen2.5-1.5B-Instruct

A small but mighty Qwen2.5 model, great for quick local inference or resource-constrained applications. (text-generation, 14.3M downloads)

Qwen/Qwen2.5-7B-Instruct

A versatile 7B Qwen2.5 instruct model, balancing performance and local deployability for many tasks. (text-generation, 13.3M downloads)

meta-llama/Llama-3.1-8B-Instruct

The latest Llama 3.1 8B instruct model from Meta, a strong contender for local reasoning tasks. (text-generation, 10.8M downloads)

pyannote/speaker-diarization-3.1

Need to identify who said what and when? This model provides state-of-the-art speaker diarization. (automatic-speech-recognition, 10.3M downloads)

Qwen/Qwen2.5-3B-Instruct

A solid 3B Qwen2.5 instruct model, offering good performance for many NLP tasks on less hardware. (text-generation, 10.1M downloads)

banned-historical-archives/banned-historical-archives

A unique image dataset for training models on historical archives, perhaps for censorship detection or analysis. (2.0M downloads)

Salesforce/wikitext

A classic language modeling dataset for training text generation or fill-mask models. (1.3M downloads)

Maynor996/img_upload

Another small image dataset; useful for testing image processing or classification models quickly. (1.1M downloads)

Tycoon AI

Automate your entire solo operation with AI agents, finally freeing up your time for real work. (444 upvotes) [producthunt]

Mintlify Workflows

Keep your knowledge base fresh automatically, so you don't have to manually update docs ever again. (306 upvotes) [producthunt]

WeWeb 3.0

Build apps with code, but keep the visual safety and speed of a no-code editor. (273 upvotes) [producthunt]

Google Antigravity 2.0

Visually build and manage complex multi-agent AI workflows right from your desktop. (265 upvotes) [producthunt]

Slideshot

Let AI agents create your product demo videos, saving hours of manual recording and editing. (196 upvotes) [producthunt]

Vivaldi 8.0

Get complete control over your app's UI with a unified look, allowing full customization flexibility. (142 upvotes) [producthunt]

Mixpanel Headless

Access product analytics programmatically, enabling AI agents and developers to build data-driven features. (130 upvotes) [producthunt]

Visual Usability Checker

Get instant AI feedback on design choices, speeding up iteration and improving user experience. (129 upvotes) [producthunt]