Founder of Junwatu. Sharing AI engineering and creative experiments.

Remote
Pinned Tweet
Little Dragon. You never know what you'll find in the woods. Made with MiniMax H3 to save credits, while Seedance 2.5 handled the more complex scenes. Created mainly with PixVerse. #PixVerseCPP @PixVerseCreator @PixVerse #PixVerse #PixMoreChallenge
8
1
21
692
There are so many ways to generate images and videos using GMI Cloud. For this one, we'll use Playground. Playground is a simple, user-friendly interface on GMI Cloud where you can experiment with AI models and generate images or videos directly, without writing any code or setting up an API. Just choose a model, enter your prompt, adjust the settings, and hit generate. It's a great starting point if you want to explore what different AI models can do, test creative ideas, or simply start generating content right away. Now, let's see how we can generate this video! ⬇️ @gmi_cloud
1
37
Kid and the Bull Created with FLUX 3 Max (T2V), then upscaled using Flux Video Upscaler.
Made with AI
6
170
Using Image to Video in FLUX 3 Max via @fal Very fast generation and only cost $0.64 for this 15s video.
Made with AI
3
223
I recently wiped my MacBook and started fresh with a clean macOS installation. And as always, the same question came up: What should I install first? So I made this video to keep track of my must-have apps, mostly so my future self doesn't have to remember them all next time. 😅 Created with HyperFrames Studio.
Made with AI
12
31
658
This prompt was exported from HyperFrames Studio, so I can't guarantee that it will produce the exact same results: > Create a 95-second English video in HyperFrames titled “Fresh macOS. 15 Tools. Let’s Build Something.” FORMAT 1920×1080, 16:9, 30 fps. A polished motion graphics tour of a developer and AI creator’s fresh Mac setup. Keep scenes, text, graphics, and audio editable. VISUAL STYLE Use the junwatu look: - Background: #090909 - Panels: #060B08 - Main text: #B7DEFF - Accent: #C0EFC7 - Secondary text: #8BAD92 - IBM Plex Mono, regular and medium weights. Use generous spacing, large headlines, rounded dark interface windows, terminal commands, symbolic tool badges, subtle gradients, and a segmented progress indicator. Place tool names and explanations on the left, animated interface demonstrations on the right. Add small, witty supporting lines. MOTION Use crisp staged reveals, smooth typography entrances, purposeful cursor movements, responsive interface animations, and quick wipes between tools. Every demonstration should show an action and its result. Keep text readable and avoid dead time. TIMELINE AND NARRATION 0–6s — Opening Show realistic first-person hands opening a MacBook on a charcoal desk. Move the camera toward its display until the screen fills the frame and transitions into the first tool. Title: “Fresh macOS. 15 Tools. Let’s Build Something.” Narration: “Fresh macOS. Fifteen tools. Let’s build something.” 6–11s — Python Animate a terminal command and an AI workflow. Narration: “Python powers many of the AI tools I use.” 11–16s — Node.js + npm Show a local web development workflow with React, Tailwind, and Three.js. Narration: “Node and npm power my local React, Tailwind, and Three.js projects.” 16–21s — Grok Show a conversational AI interface. Narration: “I access Grok through my X subscription.” Keep subscription claims specific to the creator’s account; do not imply every X plan includes every service. 21–26s — Grok Bot Animate connected services feeding a task checklist. Narration: “Grok Bot helps with routines through tools I connect.” Supporting text: “Connected services and permissions required.” Treat scheduling or wallet monitoring as configured workflows, not guaranteed built-in features. 26–31s — Antigravity CLI Show a coding agent receiving a terminal task and returning results. Narration: “Antigravity brings coding agents straight into my terminal.” 31–36s — OpenCode Show a model picker and a terminal coding session. Narration: “OpenCode lets me try models, including free options when available.” Do not promise permanently free models or describe them as mysterious. 36–47s — Cursor Give this tool a longer demonstration: select code, ask AI for an edit, inspect a diff, apply the change, then generate files. Narration: “Cursor is my IDE for vibe coding. Select code, ask AI, review changes, apply edits, and generate files.” Mention any subscription bundle only if verified for the creator’s actual plan. 47–52s — ChatGPT Animate brainstorming, research, and proofreading exchanges. Narration: “ChatGPT helps with brainstorming, quick research, and proofreading.” 52–57s — GitHub CLI Show repository and pull request commands with successful results. Narration: “GitHub CLI connects my repositories with AI tools and agents.” Present it as a command-line tool used by automation, not as an AI agent itself. 57–62s — Modal CLI Animate a laptop submitting a workload to cloud GPUs. Narration: “Modal gives my laptop access to cloud GPUs.” 62–67s — Hugging Face CLI Show model and dataset downloads with progress and completed files. Narration: “Hugging Face brings models and datasets to my Mac.” 67–72s — HyperFrames Studio Show editable video tracks, a moving playhead, and an animated preview. Narration: “HyperFrames Studio turns code into animated video.” 72–77s — DaVinci Resolve Show editing tracks, color controls, and audio meters. Narration: “DaVinci Resolve handles editing, color, sound, and the final polish.” 77–82s — Discord Show community conversations and screen sharing. Narration: “Discord connects me with the people I build with.” 82–87s — Amphetamine Show an active keep-awake session and a simple coffee-cup motif. Narration: “Amphetamine keeps my Mac awake while the work keeps running.” 87–95s — Finale Assemble all fifteen tool badges around a central MacBook, connecting them into one creative ecosystem. Main text: “CODE. CREATE. AUTOMATE. REPEAT.” Closing text: “Fresh Mac. Fresh Start.” Narration: “One connected creative toolkit. Code. Create. Automate. Repeat. Fresh Mac. Fresh Start.” AUDIO Generate a warm, friendly male English narration with clear pronunciation. Fit delivery naturally into each scene. Add understated instrumental electronic music, ducked beneath speech, with a gentle opening and closing fade. CONTENT ACCURACY Before generation, verify current product names, capabilities, and subscription eligibility using official sources. Preserve the qualifications above in both narration and on-screen text. FINISH Keep essential content inside safe margins. Check layout, readability, motion, audio, and scene transitions. Inspect representative frames and fix all errors before handing over the video.
1
1
168
Finally, P-Video-2-Pro-Avatar model is really good at creating avatars! Here are a few examples I generated using just still images. @PrunaAI
Introducing P-Video-2-Pro-Avatar, based on @MiniMax_AI H3 ! Turn a single portrait into a realistic talking video with natural expressions and accurate lip sync. And for launch, it's 50% off for the first two weeks, until November 5 at 18:00 CET! • 𝗣𝗿𝗶𝗰𝗶𝗻𝗴: $0.02/s at 480p · $0.04/s at 768p • 𝗦𝗽𝗲𝗲𝗱: ~0.97 s at 480p · ~2.2 s at 768p per second of output video • 𝗜𝗻𝗽𝘂𝘁: One portrait with a text script or uploaded audio • 𝗩𝗼𝗶𝗰𝗲𝘀: 30 voices across 10 languages • 𝗗𝘂𝗿𝗮𝘁𝗶𝗼𝗻: 1s–3min at 24 fps, in 480p or 768p • 𝗖𝗼𝗻𝘁𝗿𝗼𝗹: Voice delivery, scene motion, seed, and prompt upsampling One image, two ways to bring it to life. Use your own audio or generate speech directly from text. Available via our inference partners @replicate, @eachlabs, @Cloudflare, @scenario_gg, @wavespeed_ai, @WiroAI, Sociaro, @TellersAI, and @togethercompute. 📚 Model docs: docs.api.pruna.ai/guides/mod… 🧩 Try the model through the API or Playground: dashboard.pruna.ai/login
Made with AI
3
25
710
Bro looks like he was born in a palm oil plantation and raised by pure chaos. 😭🌴💀 I call him SAWIT MAN! 100% palm oil monster, 0% mercy! 🔥 Made with FLUX 3.
Made with AI
4
22
411
Kurang dari seminggu jalanin tim kecil pake Grok Bot dan merka uda bisa ngurangi banyak buanget kerjaan manual, jadi beneran membantu irl. Sekarang Claude ngeluarin Duet. Tak pikir-pikir ini adalah future "user interface". Kalau kalian gimana, pake Grok Bot, Muse, Dots apa VPN?
We’re opening a research preview of Claude Duet, a new app where Claude works as a proactive agent. It comes to you with what needs your attention, completes tasks you ask of it, keeps working when you step away, and checks with you before it acts on your behalf.
1
2
12
576
Carmen (Hearts2Hearts) ngaku di sela fan sign Jakarta kalau dia nggak peduli sama celah keamanan PHP, soalnya semua sistemnya udah dipindahin ke Rust sejak 2024. "Fansku nunggu aku ke Jakarta 2 bulan. Tim dev nunggu rewrite Rust udah 30 bulan, masih santai aja. Ternyata aku lebih cepet," katanya.
3
222
DΞV retweeted
No tricks. Just tracks. 🎃 Sonilo’s Halloween giveaway is here: - Free users: 1 month of Pro, free - Active paid subscribers: 100,000 bonus credits that never expire Follow @Sonilo_music, repost, & leave a comment to join. Ends Oct 31. Let’s make something spooky.
147
129
439
107,037
AI creativity, it turns out you can do literally anything you want! For this video, Seedance 2.5 still struggles a bit with the Javanese language, but visually, the result is already pretty close to what I imagined. Now, as long as you have enough tokens, the only limit is your wallet! 😂 Created with PixVerse's Agent. @PixVerse @PixVerseCreator #PixVerseCPP #PixVerse #PixMoreChallenge
Made with AI
3
9
406
This 24-second video cost me less than $1 to make! I used p-video-2-pro-reference by @PrunaAI with just 3 character reference photos to keep the same girl consistent across 7 cities: ✈ Bali → Santorini → Istanbul → Marrakech → Bangkok → Kyoto → Seoul Pretty impressive!
Made with AI
16
24
514
Devil transformation at the dinner table. #Lovart, #LovartPartner
Made with AI
4
9
181
DΞV retweeted
DeepSeek V4.1 Flash 1017 is now live on GMI! We know many of you have been waiting for this one. You can now run it with your GMI API key or access it through GMI's endpoints on OpenRouter. Happy building!
4
2
32
1,201
Eksplorasi Rumah Joglo 3D Di Indonesia itu sumber inspirasi luar biasa banyak, mulai dari rumah tradisional, pakaian adat, artefak, dan masih banyak lagi. Potensi edukasi, pelestarian budaya, sampai pariwisata itu seriusan gede banget! Dan untuk eksperimen ini, saya pakai Gemini 4 Argon (Pro subscription) lewat Antigravity untuk bikin web app eksplorasi 3D dari rumah joglo. Saya gak pake referensi apapun jadi di-handle sama Gemini. Akan lebih baik kalau misalnya ada real aset buat referensi. (Thanks to @thebuggeddev for initial the source code)
Made with AI
4
3
22
537
Kuota Gemini 4 Argon habis sekitar 40% hanya untuk bikin web app di atas.
74