FramePack: Generate Video Forever

[NVIDIA ONLY] The creator of ControlNet just dropped a new approach to generating videos in a PROGRESSIVE way--you can generate loooong videos with LOW VRAM (just 6GB)

@SUP3RMASS1VE wrote a 1-click launcher to run the Gradio app on your PC! Get it on pinokio.computer

Just ran a real world experiment in chatgpt. it's SHOCKING how well it works. Here's how it went:

Step 1. "i have a cup of olive oil and a cup of water. i mixed them in a transparent cup. generate an image of what it looks like now" Step 2. "now i've added some boba, generate an image of the result."

Run ZONOS Locally

ZONOS, the new SOTA Open Source Voice Cloning TTS, is here.

I've managed to write a 1-click launcher for Zonos that works on Mac, Windows, and Linux (ALL platforms!)

Here's me cloning Peter Griffin's voice on my Mac. Available for 1-click launch on pinokio.computer

Run Deep Research on your PC Locally with 1 Click.

Browser-use really works! I asked:

"Compose a report on pinokio, what kind of unique value you can get you can't get with anything else."

Watch as it crawls around the web with an actual browser. And you can do it too! First, I wrote a 1-click launcher so you can easily install and run browser-use Web UI

https://x.com/cocktailpeanut/status/1886451071356391815

Top Local & Open Source AIs Launched in 2024

Open source AI is great because we can simply download and run them locally with no internet connection, privately, and for free.

Here is the year 2024 in review. A subjective list of top open source AIs launched in 2024 (Thread) 1. Top Voice Cloning AI: F5-TTS

F5-TTS came out of nowhere and just dominated. Yes there have been open source voice clone AI models before, but none this good.

Other models stepped up since then (EX: Fish), but F5 was the first to make the quantum leap

https://x.com/cocktailpeanut/status/1845542481360031982

MMAudio Comes to Mac!

MMAudio is one of the BEST AI models that came out this year, period.

Just give it a video clip to get audio clips that match the video. It's magical.

And now Mac users can use it too. The MMAudio gradio app runs on Macs now (I implemented it! 😎) Get it with 1-click on pinokio.computer

Pyramid Flow over FLUX

Pyramid flow is a video gen AI that's been around a bit, but the original version was bad (based on SD3).

But they retrained the model using FLUX. Now it's REALLY good. Both txt2vid and img2vid. Even works on Macs! Gradio App 1-Click Launcher is here. Again, works on ALL platforms, even Macs! (Tested on my Mac M1 Max 64G). Probably the first good quality video generator with Mac support.

Also, on GPU machines it uses CPU offloading, so all you need is less than 8GB VRAM to use this.

Available on pinokio.computer

100% Local LLM Coding with Bolt + Ollama + Qwen

@Alibaba_Qwen just dropped their qwen2.5-coder-30B and it's on @ollama. Finally a model good enough to use instead of gpt-4o or claude.

Now with Qwen, Bolt perfectly displays the app preview with Ollama.

Run it with 1 click. 1-click launcher on pinokio.computer

Use GPT-4o for FREE with Bolt

Crazy update just dropped. Don't want to pay for LLM API keys just to use Bolt?

Now the Bolt 1-click launcher lets you use GPT-4o and GPT-4o-mini completely for FREE, powered by Github and @LiteLLM - watch me get started in 1 minute: Getting started:

1. New users: you can install with 1-click on
2. Existing users: this update has some critical changes, so just to avoid any confusion or issues, please just delete the existing one and reinstall from scratch.pinokio.computer

Run Bolt Locally, with 1-Click.

Mind blown. All non-technical people building apps with AI on a LOCAL machine should try this.

I wrote a 1-click launcher that supports @OpenAI, @AnthropicAI, @GeminiApp, @OpenRouterAI, @GroqInc, @MistralAI and even local LLMs via @ollama Available on right now.

On ALL platforms, regardless of GPU. pinokio.computer

When Should You Use Local AI?

I myself use ChatGPT and Claude and they are super helpful. When it comes to LLMs I do recommend using them at this point in time.

So when SHOULD you use AI that's running locally on your machine?

1. Privacy
2. Free
3. Censorship-proof 1. Privacy

As someone who values privacy A LOT, I am very cautious about sending over my info to a 3rd party service.

I don't think I will send over a live stream of screenshots of my computer screen to a 3rd party service. For these cases local makes sense.

F5-TTS Finetune Web UI

The E2-F5-TTS project has a huge update. Now anyone can easily finetune the model with their own audio using a Gradio app!

It automatically transcribes the audio clips using whipser, and seems to work for most languages. Here's a Korean example. This new Finetuning UI is now included in the official Github repository, so all you need to do is pull in the updates in Pinokio (or install fresh)

https://x.com/cocktailpeanut/status/1845542481360031982

Elevenlabs-level TTS on your laptop.

I'm always skeptical about new AI models hyped up sounding too good to be true. But this...is crazy good.

And now, you can run the gradio app by @realmrfakename on your laptop with 1 click.

Meet @elonmusk, from Silicon Valley. Available on pinokio.computer

Facefusion 3.0 is here!

The 100% open source face swap app powered by Gradio, launched v3.0, supporting all kinds of new features including:

- Age Modifier
- Face Editor (via LivePortrait)
- Workflows! (Job Queue System)
- And more You can install with one click via pinokio.computer

Video-to-Video for CogVideo

CogVideo video-to-video diffusers pipeline just dropped---it lets you take any video and turn it into another video.

So I've added a "video-to-video" tab to the CogVideo Gradio app.

Example: Turn a car driving video into a video game version. Run it with 1 click

ps.

If you already installed the previous version, just update to the latest by clicking the "update" button on the sidebar. If that doesn't work, just reinstall from scratch.

https://x.com/cocktailpeanut/status/1830266948808040831

ComfyUI Launcher

- ComfyUI has shipped a new frontend architecture.
- FLUX is now the go-to model for image generation.

So I wrote a new 1-click launcher for the new ComfyUI, optimized for FLUX.

No need to download custom workflows or models, all automatically taken care of. The launcher script available on pinokio.computer

Replace ChatGPT with Llama3.1, instantly.

It's super easy to replace ChatGPT with a private & local LLM (Llama3.1), now better than centralized LLMs, powered by @ollama + @OpenWebUI.

Step 1. Get OpenWebUI 1 Click Launcher
Step 2. Download Llama3.1 from Ollama
Step 3. No step 3! Step 1. Install OpenWebUI

Get the latest version of @openwebui inside

- If you're new, just get it with one click.
- If you already have OpenWebUI installed, delete it, and install this new one from scratch.

You will need Pinokio v2.0.pinokio.computer

Introducing Pinokio 2.0

A complete rethinking of how offline web apps & AI apps should work.

- Invisible: pinokio-powered offline web apps should be indistinguishable from normal websites
- Frictionless: 0-click launch
- Shareable: turn local machine into a public web service Update to the latest version (2.0.2 as of now) here: program.pinokio.computer/#/?id=install

Gligen GUI 1 Click Launcher

Simply draw regions to turn your imagination into images.

And now, you can run Gligen GUI on your local computer with one click. Works on ALL OS (Windows, Linux, Mac). To install, you MUST upgrade to Pinokio 1.2.

Again, UPDATE TO PINKIO 1.2

After upgrading, install Gligen from the discover page.docs.pinokio.computer/download/

Vid2Pose 1 Click Launcher

A dead simple Gradio app that lets you convert ANY video to OpenPose & DWPose.

And now anyone can install and run it on their computer locally, with 1-click.

Use this to generate DWPose videos to use in AnimateAnyone.

Supports: Mac, Linux, Windows You can find it on pinokio.computer

Introducing DEUS, a Realtime Image Generation Engine.

1. Realtime Native
Designed from scratch with a "everything is realtime" assumption

2. Simple
If you can draw, and google, you know how to use DEUS

3. Intuitive
Create exactly what you imagine without special knowledge
GitHub

github.com/cocktailpeanut…