Daily Digest – 2026-06-23

JohnBra/vite-web-extension

A solid base for browser extensions with modern tech, so you can build features, not boilerplate. (TypeScript, 1.2k★)

★ 1,198

mugnavo/tanstarter

Jumpstart your SaaS with this template: auth, ORM, and UI included so you ship faster. (TypeScript, 1.1k★)

★ 1,133

guigrpa/docx-templates

Generate Word docs from templates, handy for reports or custom client exports without fighting Word's API. (TypeScript, 1.1k★)

★ 1,072

Skolaczk/next-starter

This Next.js starter has auth, payments, and testing baked in; basically a jumpstart button for your SaaS. (TypeScript, 987★)

★ 987

nateraw/stable-diffusion-videos

Turn text prompts into cool Stable Diffusion videos; great for generating unique visual content on a budget. (Python, 4.7k★)

★ 4,684

vllm-project/vllm-omni

Efficiently serve multimodal AI models, letting you run complex AI faster for less cash. (Python, 4.6k★)

★ 4,580

antgroup/echomimic_v2

Generate human animations from audio; useful for creating engaging talking heads without complex setups. (Python, 4.6k★)

★ 4,551

Tencent-Hunyuan/HunyuanVideo-1.5

Generate videos from text or images with this lightweight model, ideal for quick content creation. (Python, 4.4k★)

★ 4,401

openai/gpt-image-2

OpenAI's advanced image generation, offering precise text prompts and editing for detailed visual content. (8.8M runs)

google/gemini-3-flash

Gemini 3 Flash offers speed and advanced intelligence, perfect for rapid conversational AI or search features. (5.0M runs)

google/upscaler

Upscale images without losing quality, useful for sprucing up old assets or low-res client images. (673K runs)

ultralytics/yolo11n

YOLO11n for real-time object detection, a lightweight and fast model for vision tasks. (43K runs)

ultralytics/yolov8s-worldv2

YOLOv8s Worldv2 offers real-time open-vocabulary object detection; ideal for dynamic visual analysis. (16K runs)

granite3-dense

IBM's Granite models for RAG and code tasks, useful for local tool-based AI without cloud dependencies. (939.1K pulls)

granite3.1-dense

IBM's Granite 3.1 models offer improved speed and performance for local text generation and RAG. (936.7K pulls)

dolphincoder

An uncensored coding LLM (7B/15B) based on StarCoder2, useful for unrestricted local code generation. (924.6K pulls)

wizard-math

A specialized LLM for math and logic, ideal for complex problem-solving where accuracy matters. (921.9K pulls)

llama3-gradient

Llama-3 8B with a massive 1M+ context window, great for analyzing huge documents locally. (921K pulls)

@appium/tsconfig

Standardized TypeScript config for Appium projects, making cross-platform test automation setups easier. (1.6M/wk)

@appium/base-plugin

A foundational plugin for extending Appium, letting you customize mobile test automation with less hassle. (986K/wk)

@clawdbot/lobster

Build robust AI agent workflows with approval gates, ensuring your automated tasks don't go rogue. (544K/wk)

@genkit-ai/google-genai

A Genkit plugin to easily integrate Google's AI models, especially Gemini, into your Genkit flows. (142K/wk)

genkit-cli

CLI for the Genkit AI framework, letting you manage and test your generative AI projects efficiently. (106K/wk)

playwright

Automate browser tasks with ease; perfect for scraping, testing, or building bot workflows. (20.8M recent downloads)

akhaliq/anycoder

A Dockerized coding assistant, offering quick access to AI-powered code generation right in your browser. (docker, 3290 likes)

zerogpu-aoti/wan2-2-fp8da-aoti-faster

A Gradio demo showcasing faster inference with specialized models, useful for benchmarking your own AI projects. (gradio, 3245 likes)

HuggingFaceTB/smol-training-playbook

A playbook for small-scale model training, making AI experimentation and learning more accessible. (docker, 3214 likes)

mukaist/Midjourney

An unofficial Gradio interface for Midjourney, providing a free way to try text-to-image generation. (gradio, 2980 likes)

pharmapsychotic/CLIP-Interrogator

Describe images with text using CLIP, great for getting prompts for AI art or understanding visuals. (gradio, 2972 likes)

intfloat/multilingual-e5-large

A large multilingual embedding model for semantic search or text similarity across many languages. (feature-extraction, 8.1M downloads)

openai/gpt-oss-20b

A substantial open-source GPT model for text generation, letting you self-host a powerful conversational AI. (text-generation, 7.0M downloads)

Qwen/Qwen3-4B-Instruct-2507

A compact, instruction-following Qwen3 model, great for local or low-resource text generation tasks. (text-generation, 5.6M downloads)

dima806/fairface_age_image_detection

Detect age in images, useful for content moderation or audience analysis without paying for an API. (image-classification, 5.5M downloads)

antirez/deepseek-v4-gguf

A quantized Deepseek-v4 model in GGUF format, perfect for running powerful LLMs efficiently on your hardware. (text-generation, 5.2M downloads)

ayuo/hd_tmp

Just a temporary dataset, probably for internal use; move along, nothing to see here. (1.5M downloads)

hf-doc-build/doc-build-dev

A development dataset for documentation builds, handy for testing tools that process docs. (865K downloads)

princeton-nlp/SWE-bench_Verified

A verified benchmark dataset for software engineering tasks, ideal for evaluating your code generation AI. (699K downloads)

MD+HTML Reader

Review AI-generated Markdown and HTML effectively, saving you time fixing bot mistakes. (116 upvotes) [producthunt]

OnBrand by SlideSpeak

Design context for AI agents, so your bots actually understand what they're doing. (108 upvotes) [producthunt]

Agentic Document Extraction

Turns documents into computable data, making any PDF or text searchable and actionable. (101 upvotes) [producthunt]

AlgoFly AI

An all-in-one platform for building and deploying vision AI, simplifying complex computer vision projects. (95 upvotes) [producthunt]